Game analysis - Replay #14886292

zappazapper

DISCLAIMER: Although I'm already aware that in general, even the most "toxic" players on FAF want everybody to be better at this game and are very helpful in pointing people to the resources they need to improve, I feel it's necessary to point out that my reasons for doing a game analysis are purely for my own education and not as some kind of attempt to flex. My 1v1 rating was -72 when the game started and -45 when it ended. It's been as low as -376. In no universe is there a reason for me to flex, so relax, I know I'm terrible. But I want to get better so I'm using this primarily as a way to work through my own thoughts on what I did right and what I did wrong, and if anyone has any helpful comments I'll be happy to read them.

Map: Festea VII

Less than a week ago I was so frustrated with this map that I made a point of saying so on #aeolus. Nobody cared, of course. I realized that the reason I said that was that I was hoping someone there would tell me I need to study the map then. So I studied the map. What I realized is that I need to get better at reading terrain. This is actually a pretty good map for someone of my skill level, because there are only 3 points of entry by land, which should be easily enough defended.

Opponent: gliffer_slash (0)

I assume that zero rated players are new to FAF but not necessarily new to SupCom. I've been gutted by a number of zero rated players, hence the -376 and the slow crawl out of the basement. I've also played a number of players with lower ratings than me that were better, and >0 rated players who were worse, so I've learned not to sit back and relax, no matter what the rating is. Really, gliffer_slash is not a bad player compared to me and it was not a cakewalk, but he did make some errors that allowed me to win.

Build order:

Like many new players, I struggle with build order. I especially struggle because I'm trying to incorporate reclaim into my build orders, and that takes a lot of attention/APM that is the most scarce resource in every game I play. Also, in the game previous to this one I was rushed by bombers and realized that I had to find a way to get some interceptors into the air quicker, and as Crag Dunes was the map that this happened on, that was the map I used to work that out. Unfortunately, once the game started and I hit CTRL-SHIFT to take a look at the reclaim situation, I realized that this map doesn't have a ton of concentrated mass lying around, so I kind of abandoned that part of my eco strategy for a bit and just sent my engineers out to build mex's.

The build order I'm using is as follow:
Land Factory with commander, building engie x 1, LAB x 2, then engie x 5
2 pgens with commander
2 mex's with commander
1 pgen with commander
2 mex's with commander
1 pgen with commander

The commander-built pgens were built all next to the land factory.
The first engie that comes out I had build 4 pgens perpendicular to the ones the commander built, and then an air factory such that it received an adjacency bonus from all 8 pgens. Once the commander was done its build order I had it assist the engineer on the remaining pgens and air factory.

The LABs I had run down either side of the map to hunt for engineers. The LAB on the right unsuspectingly encountered my opponent's commander (blap), and the one of the left ended up being reclaimed by the engineer he encountered. I didn't even know that was a thing. Moving on.

As I said, my developments in learning proper build orders thus far have depended on a fair amount of mass reclaim available, and when I realized that this was not a feature of this map, I panicked and just sent the 5 engineers on mex building missions. As a result of this and my attempt to build pgens to prevent power locking upon building an air factory, I did mass lock for a bit. Fortunately, in my confusion I ended up sending the engie that was building those pgens on a mex building mission, which relieved the lock enough that my commander could finish building the 4 mex's at the starting location. Then I took another engineer and continued to build the pgens I needed and the air factory.

On the subject of engineers - I really don't know what an appropriate number of engineers is, and when. When I played vanilla SupCom single-player I would always have a factory spamming T1 engies (most build power / mass cost) on a control number and I would just move the whole mess wherever I needed build power. When I started playing FAF I had to adjust to engie-mod, where it makes more sense to build factories to get build power. As a result I scaled back my production of engies for the sole purpose of early base expansion, but then there are always times when a big cloud of engies might be useful, so that's something I need to investigate further. Anyway, that's the reason I only built 6 engineers out of the gate and put that first factory on army production.

Here's the unit composition I've been using and my reasons for it:
First off, I've been reading https://forum.faforever.com/topic/766/ladder-1v1-beginner-intermediate-and-advanced-topics-by-arma473 like it's a bible, and one of the things it says is a tank to artie ratio of 5:1 is ideal. I have read other resources that suggest having a scout bike in the mix, MAA in the mix, and an engie in the mix. I'm sure there are better unit compositions for different situations but I've reached the conclusion that for someone of my skill level, as varied a unit composition as possible is probably the safest bet. So,

Tank (tanks provide hit-points, and thus survivability, and a DPS soak, so you need lots)
Scout Bike (provides short range intel and extends weapons range)
Tank
Mobile Anti-air (nobody's winning wars with T1 MAA, but they do soften up bombers so that they do less damage before an intie can deal with them)
Tank
Artillery (it took me a while to realize that despite the >2x DPS of an artie vs a tank and somewhat similar HP, artillery is only useful against stationary targets, and without a bunch of tanks to stop an opposing force from moving forward, arties are useless, so only 1 for every 5 tanks)
Tank
LAB (almost same DPS as a tank, but far less HP... my rationale here is that I want some speed on the field for quick, short range raids on mex's)
Tank
Engie (for building mex's as territory is gained, for building point defense/static AA to hold a position, for building radars, for building forward factories)
repeat

Now, I also realize that this unit composition might be ideal for the first run, and then it might be a good idea going forward to reduce the concentration of certain units, for example, you probably don't need 20 scout bikes mixed in with a blob of 100 tanks. At this point in my "career", I don't have the attention/APM to be that exacting with such things, and at any rate, scout bikes are the cheapest things you can build besides walls, so really, I think I'm getting a good mass cost on attention/APM.

Out of the first air factory:

Bomber
Air Scout
Bomber

Bomber first because they're slow, and by the time it's near its target and likely to need intelligence, the scout is probably already built (it's cheap) and ahead of the bomber. Second bomber because either a) the first bomber got killed but the target has been softened, or b) the first bomber isn't dead because the opponent hasn't got their AA out yet, and I should jump on the opportunity to do some damage.

At that point it's a decision on how to continue with that air factory. If he's also sending bombers my way then I need to start making inties. To be honest, I don't even know if he was making bombers, but I got fucked so hard in my last game that I just started making inties.

3 Inties
1 Air Scout
repeat

Now, the reason for the star-shaped patrol is because it's another good price to pay for attention/APM. It's not something a 2k player is ever going to do, and I realize that air combat is something I will have to get better at going forward. But also, the opponents I'm facing at this point don't really require better air play, so it's fine for now. Also, it does give me a couple advantages in return for what I lose by not micro'ing it a bit more - the 5 point star shape gives good coverage, both on the interior and exterior of my territory. Also, it didn't happen in this game but I do often select all my inties with a hotkey and give them specific targets, and the advatage of having them all flying around in a cloud is that on average, every intie is just as close to anywhere else on the map than every other intie, as opposed to guessing on a good place to stage them where they'll be in a good place to react but also safe.

EDIT: So as I'm typing this, I'm watching the replay, and I realized that I fucked up. I didn't build the second group of pgens so that they would be adjacent to the air factory, i built them on the land factory. Like I said, I panicked a bit.

Next, with that first land factory producing units to defend the east point of entry, I proceeded to build factories to cover the west and south points of entry. Same unit composition. Then I built 3 more factories to assist the factories building units for the 3 points of entry. At that point I started power stalling, so I started building pgens with my commander and reclaiming trees with idle engineers. Once my power situation was dealt with, I built another 2 air factories to spam out bombers. An argument could be made that I should have built more land factories (mass was overflowing), but in previous games I had abandoned the bomber campaign as soon as my opponent had static AA on the field, and I wanted to see if I could overcome it, and also my opponent was not making any sort of push via land so I figured I had some time. Ultimately, my opponent built a lot of static AA and bombers proved ineffective at clearing a path to hit his hydro. At that point, I had started building land factories to build attacking armies for the middle and eastern passages, which caused a power stall, then a mass stall, so I put my intie spam and one bomber factory on pause.

Once the first units of my middle passage attack force reached their rally point at his northern entry point, he attacked with an army that had some T2 tanks and began to push through the passage. At that point I should have thought about starting to upgrade to T2 myself but I was in the throughs of a mass stall, and despite not having the intelligence to make such a determination (no radars built the whole game... derp), I figured I had the T1 spam to deal with it if he got as far as my southern entry point. Fortunately I was right, even if it required the assistance of the army defending my western entry point. Also, again without intelligence to justify it, I figured that he had committed a good percentage of his forces to this push, and despite the fact that I might have been leaving myself open to a counter-attack that I couldn't defend, I pushed down the east passage with the army that was defending my eastern entry point. Once the battle in the middle passage had concluded, I realized he was pushing up the western passage, so I desperately attempted to run half of the army in the middle passage to help cover the vulnerable force I had at the western entry point. Fortunately I managed to contain his advance.

My attack on his eastern point of entry was successful enough, in that I had won a battle of attrition and was able to start softening up his structures, but I didn't have the units to push any further from the east. I can't talk enough about how disappointed I am in my lack of intel, so I won't bother. Regardless, I had the FEELING that I had him on his heels, so I pushed forward with whatever I had, which in this case was a limited force at his northern entry point. Again, I was fortunate that I had guessed correctly and that he had committed to defending my eastern attack. I managed to take out some T2 mex's and a land factory. Also, by that point his AA was less of a factor and I was able to make use of some bombers I had lying in wait.

In terms of land forces, again with a complete lack of intel that I'm embarrassed to fess up to, I decided to let my armies collect at their rally points at his north and east entrance points, while at the same time pushing forward in the west. In truth, I didn't do a very good job on the west side of the map: I was fortunate that he valued the west side of the map as poorly as I did. I could have easily won the battle on the west side, if I had built units as I did in the middle and east, but because of my own miscalculation/laziness/lack of attention, my push on the west side was limited to a purely diversionary tactic. It worked, although not perfectly. My waypoint was too far forward, which caused me to lose units unnecessarily. In a rare moment when I had the attention to pay towards this maneuver, I pulled the waypoint back to allow my units to collect into some semblance of a formidable force.

As insignificant as it might be in the whole scheme of things, I think I might be the most proud of this particular action. I have, for so long, depended on the automation of things in this game. There's a reason why my 1v1 rating was -375, and is currently -45. I don't think I'll ever get to 2k+, but I think I'm allowed to take a certain amount of limited pride, as a sub-zero rated player, in the realization that I'm not HITLER... I have to give my troops reasonable orders... I'm ultimately responsible for their well-being... It might just be a game, but the health of my units = the health of my strategy/tactics...

At any rate, my western push was more successful than I had hoped, considering my limited resources. Again, my lack of intel prevented me from knowing things that are clear in the replay, and I endeavor to take no credit from "hunches". I need to do a better job of radar and scouting. But I would be lying if I said that I didn't think that this push was a worthy endeavor, on a hunch, and luckily I was right: my opponent did move his commander west to try and minimize damage.

At this point, in an attempt to avoid a full analysis of my opponent's skill and strategy, which I am clearly not qualified to provide, I will point out that I have learned a lot by reading what the high-level players of this game have suggested I read.

Again...
https://forum.faforever.com/topic/766/ladder-1v1-beginner-intermediate-and-advanced-topics-by-arma473

There's a lot here, but if I can give any advice from one noob to another... you don't need Tech2, and you certainly don't need Tech3... at some point, we all will need to figure out how to upgrade to higher tech levels to compete, but if you're playing ME, it means you're terrible, and if you're terrible, it means you haven't figured out Tech1...

The secret to Tech1 = spam... spam land factories... they're actually pretty cheap, and you can pause them if you get mass locked... and from them, spam T1 units... just spam tanks if you don't know what else to do... I'm not smart/fast enough to realize that you don't have a good mixture of mobile AA... and I'm too stupid to attack your army with bombers anyway... I'm usually more focused on taking out your power, which can easily be solved with, what, 2 static AA?

The point is, my opponent's solution in this game, even though Tech2 didn't help him, was to upgrade to Tech3... That's when I knew I had him...

Tech2 is a great compliment to a GREAT Tech1 army... Tech2 gatling bots add DPS and range that will blap a T1 army into non-existence... T2 tanks are the most efficient source of hit points, by mass cost, of any unit... T2 Missile WILL get rid of pesky T1 point defense... but a purely T2 army is useless, because it's almost impossible to manufacture T2 units as fast T1 spam... I haven't done the calculations on the build time on a T2 tank vs a T1 tank, regardless of resource cost, but I suspect it's not good...

and T3 is much the same... ya, T3 bots are great, but if they're the sole target of a simple T1 army consisting of 20 or so tanks, they're dead...

And so that's what happened... my opponent made a last ditch effort to produce T3 bots... they overkilled a few T1 tanks and then I skullfucked his commander... one could argue that I had already won the game minutes ago, but he probably could have prolonged it by spamming T1 at that point, and he certainly could have made it very difficult, if not impossible, by sticking with T1 when he went to T2 in the first place...

Things I Did Right:

There aren't many. I think sticking with Tech1 is the only thing I did right.

Things I Did Wrong:

There are many. Too many to bother. The main things I don't like are all related to eco:

I didn't build mex's as I acquired territory in the middle
I didn't bother with transports to get engies to mass points that weren't pathable
I didn't upgrade T1 mex's to T2 because i was either mass locked or attention locked

All in all, I think this is one of the best games I've played, but I'm also concerned for the future. I know that with every increase in rating comes more difficult opponents. I don't know if I have the real-world lifestyle that allows me to keep improving. I know that I thoroughly love this game and have the time to consider the meta, but I don't know if I have the time to do the in-game work to apply the meta properly.

But I love the fact that my favorite game has a META. This is such a hard game. Yesterday I wanted to quit forever because I got gutted by a bomber rush. I love that I wanted to quit yesterday and now I'm here writing a novel on my game with a legit zero-rated noob.

I'm having so much fun. Consider that when you make decisions for the future of FAF.

CheeseBerry

Very intriguing read, love it!

I don't quite agree with some of your analysis but you already knew that it's not perfect, so who cares.

As long as you keep going at it with a mindset like this, you are bound to improve. So keep up the good work!

Zokora

I would not invest too much time into those game analyses until you hit 1200+. Time would be better invested for the next game until you have some basic routines in your blood. Your mindset is good and you have a certain solid BO, you just need to get more game experience.

archsimkat

@zokora said in Game analysis - Replay #14886292:

I would not invest too much time into those game analyses until you hit 1200+. Time would be better invested for the next game until you have some basic routines in your blood. Your mindset is good and you have a certain solid BO, you just need to get more game experience.

This is completely incorrect. You need to do analysis of your replays regardless of your current level, and I would argue it's especially important at low levels (0-1000), where there are likely large amounts of low hanging fruit, ripe for improvement. Getting "basic routines in your blood" and not analyzing your own gameplay while you are at 0 ladder rating will serve only to ingrain the bad habits you currently have.

Turinturambar

@zokora said in Game analysis - Replay #14886292:

I would not invest too much time into those game analyses until you hit 1200+. Time would be better invested for the next game until you have some basic routines in your blood. Your mindset is good and you have a certain solid BO, you just need to get more game experience.

How do you want to check if you actually improved on the basics as you aim to improve on? Apart from finding new issues/new tactics in replays, watching replays can also just serve as a check how well your doing on your objectives e.g. how much e/mass did you actually stall/overflow and for how long? how many factories were idle and for how long? how fast did you actually expand? how good was your agression really? how good was your scouting/intel? how many tanks did you suicide for no reason?
While playing you will likely not notice how much you failed/not failed to fullfill your improvement goals.

biass

Putting in the effort to analyse your games is already a show of promise that the majority of newer (or, bad) players lack, so i'll try to give you some good advice here. I've watched your game and read your analysis, and there are two major points of interest that I felt needed commenting on.

First:

zappazapper said in Game analysis - Replay #14886292:

because it's another good price to pay for attention/APM. It's not something a 2k player is ever going to do, and I realize that air combat is something I will have to get better at going forward.

zappazapper said in Game analysis - Replay #14886292:

at some point, we all will need to figure out how to upgrade to higher tech levels to compete

zappazapper said in Game analysis - Replay #14886292:

At this point in my "career", I don't have the attention/APM to be that exacting with such things

I'm not sure if it's a good idea to justify bad play to yourself in a developmental stage. I don't think it's a good idea to justify anything to yourself at all, you're trying to learn and thus you must accept your mistakes so you can correct them.

Higher rated players are not doing anything particularly different that is out of your reach, they've just committed the correct things into muscle memory. If you allow yourself to continue to do something badly by justifying it with talk about your APM level and your current skill, you'll form a bad habit that will take a long time to revert; If reverting it is even possible to begin with.

After all, you'll default to your training under any form of pressure, and you'll be stuck doing something terrible like the air patrol when a higher rated player appears, or you get a new map you don't know.
I've trained new players for many years here now, and I would be confident in claiming that most of that training was beating bad habits out of people. It doesnt matter if you spend some time now doing the right thing poorly, it's much faster than trying to un-learn something later after you create a habit.

The very basic of air micro is that planes cannot shoot backwards, so it's all based around getting behind your enemy and shooting him down for free. Don't you think that would be a bit easy to do if your enemy is flying his planes around in a preset path and not reacting to you? Air patrol has negative consequences from something as small as no fuel to something as large as a complete wipe. You were not punished here because your enemy didn't even bother to make any planes to try and fight you, but be careful to not draw conclusions. You WILL be punished for doing this before you get as high as a paltry 600 rating. Learn the correct basics now.

zappazapper said in Game analysis - Replay #14886292:

There's a lot here, but if I can give any advice from one noob to another... you don't need Tech2, and you certainly don't need Tech3

You've drawn this conculsion without looking at any of the factors that might have caused him to lose here. This is a game focused around macromanagement, and it seems you've only looked at micromanagement related issues.

There is more behind why he died while he made tech, and why you've made an incorrect conclusion.

Your enemy is playing aeon, who historically have a very powerful T1 phase but a weaker T2 phase.
Your enemy made blazes, which are weaker tanks on comparison, quite possibly still the worst.
Your enemy was behind you on total mass by roughly 20k, which is roughly the price of a T4 Spiderbot.
Your enemy was overflowing mass, which made that disparity even more drastic.
Your enemy had no buildpower, so even if he stayed at t1, he could only make tanks at under half the rate.
Your enemy suicided 2 T3 bots into an army of 200 tanks, where he should be kiting you.

When you look at your games, there are a couple things for you to pay attention to in order to get the most out of it. First would be your factory counts, because if you make the same number every game, you'll win/lose depending on how much your enemy makes.

Second is your total mass collected, which you might need the Supreme Scoreboard mod to see. It's not the end all mark of why you win/lose games, but it's a stat that will help you diagnose your games in 99 percent of cases.

Third would be your reclaim count. You'll progress faster if you can quickly learn to suck up all the reclaim you possibly can and then spend it. Compare your total reclaim to the reclaim of your enemy and you can figure out why you beat/get beat by people.

You shouldn't focus on your army movements or stuff like that so early on, and its also why strange stuff like adding engineers to your army or whatever also don't matter. The bigger picture stuff is more important.

I'll stop there, one last thing though, our learning content is all made by individuals and thus I can't assure you the quality of information in each post. I could go over Arma's post you linked and point out about 100 different things I think are wrong, but someone could come to anything I make and easily do the same. Don't treat any one material as gospel; not only will you be a blasphemer, but you'll hinder your learning.

I have more I could rip you on but it's best if you don't try to internalise everything at once. Incorporate new things into your play and then come back after you do, and make sure not to force yourself into bad habits. Good luck,

zappazapper

@biass thank you for devoting some of your time and brain power to commenting on my post. This is a huge part of what I love about this game, that the better players also love this game so much that they're willing to help their potential competition improve, just because they also want to still be challenged by this game.

I have a few comments responses to a few of the things I said, and I want to make sure that you know that it's not out of a lack of appreciation for the insights you've provided to me. Despite all the talk about attention/APM, this is a very cerebral game and I think that you'll appreciate that I almost enjoy thinking and talking about the game as much as I enjoy playing (at this point, maybe more, because playing is still often quite frustrating). So when I say "ya but", please don't assume that I'm being dismissive and not considering your arguments: I'm only trying to continue having the discussion because it's both enjoyable and helpful.

On the subject of air patrols / attention / APM... I've said in previous posts how this game in some ways reminds me of another passion of mine: music. In the sense that there is a skill component (APM) and a theoretical component. What's different about playing this game and playing a musical instrument is that playing this game for the first time is a little like picking up a guitar for the very first time and trying to learn "Flight of the Bumblebee". You really should just try and learn "Louie Louie" first or some other dumb schlockey song that you'll never want to be caught dead playing in front of humans (I would) because you're not ready to attempt "Flight of the Bumblebee", just like I'm not really ready to learn how to micro Inties at this point. You're right, patrolling Inties is the wrong way to play, but if I want an opportunity to learn the things that are critically important for me to learn at this early stage, I have to put something out there because otherwise I'll get stepped on, and if I focus on microing Inties right now, I won't be focusing on what I should be focusing on, which is paying for them in the first place. And yes, I understand that all things in this game are related, that proper air play IS a critical component of an economy, in the sense that I'm not concerned with acquiring mass so I can build big shiny things, I'm concerned with acquiring MORE mass than my opponent so that I can beat him with whatever I do have, even if it's just LABs. So I do get it, and I'm not saying that you're wrong. I'm only saying that in order for me to learn, I have to get in a game, and it's unrealistic to expect to be able to improve on every aspect of my game, in every game, so I'm focusing on the things that I think are more important, like building T1 mex's and spending my mass. I think I'm getting pretty good at that, and the more I commit that to muscle memory, the more I'll have time to do things like micro air.

Concerning my comments on tech levels... I think I was making the point that it's a mistake to assume that higher tech means "better", and that if you're losing battles at T1, the solution is to upgrade to T2. It often is, but it's often not. Like you said, he had no build power and was overflowing mass even when he was behind me by 20k. Maybe if he had built more T1 factories and found out how to hit me where I was weakest, he might have been able to catch up with me on mass acquisition, or even surpass me and build a T2 or T3 army that would make short work of my T1 spam. Acquiring mass via map control and spending it on T1 units has been the single biggest change to my game from my days of playing vanilla SupCom vs AI, addicted to the pause button and turtling on 4 mex's until I could afford to build a few T3 units. So again, you're not wrong that I'M wrong, I'm just trying to relate something to new players that really helped me to correct an approach to the game that I had that was just absolutely wrong, and that I see all the time among the low-rated players I play.

Factory count - Arma's "bible" says a 2:1 factory to mex ratio is a good place to start. I don't just blindly try and build factories to satisfy that as a rule, but at the same time it did introduce me to the idea of spending my mass.

Supreme Scoreboard - the problem with Supreme Scoreboard is that the unit count and unit limit seem to be broken. I've inquired and whoever I spoke to seems to agree that it's not working properly. Maybe I need to look into that more. There is clearly some very useful information there.

Yes, reclaim is definitely one of those things I'm trying to figure out. I don't think I'm doing it right. Part of it is the mechanics of factory attack-move orders, the speed at which engineers come out of the factories, the fact that I get distracted by something else and then before you know it I have 15 engineers all going to the same place that doesn't have that much mass to reclaim. And as I said, I hadn't played this map in a little while, since I really started to reclaim more, and so I kind of panicked when there didn't seem to be a ton of mass reclaim around. I can honestly say that reclaim is at the very top of the list of things I would like to improve on.

"You shouldn't focus on your army movements or stuff like that so early on"
Mmm, that's a little bit of a contradiction there. If I don't need to focus on army movements at this point, why do I have to focus on micro'ing air?

"and its also why strange stuff like adding engineers to your army or whatever also don't matter"
I admit that the exact ratio of tanks to anything else could be refined, but I would argue that, in the right proportions, spending mass on things other than tanks really is just like spending mass on tanks. Scout bikes increase range, which will theoretically cause me to lose less tanks. MAA softens up bombers and gunships while my Inties are on their way to deal with them, which will theoretically cause me to lose less tanks. Artillery adds range and DPS, which will theoretically cause me to lose less tanks. LABs add speed, which allows me to quickly raid unprotected mex's, reducing the amount of mass my opponent can spend on tanks, which will theoretically cause me to lose less tanks. And engineers can reclaim mass, build defensive structures, and will likely be closer to a recently raided mex or a good place to build some forward unit production, which will all theoretically cause me to lose less tanks. A tank that didn't die because another type of unit gave me a slight advantage in one aspect over an opposing force of pure tanks is a tank that I don't have to build again.

"Don't treat any one material as gospel" - oh, for sure, I read lots of things. It's just that Arma's guide is kind of written like a religious text. It's long and it's the kind of thing you can go over certain things over and over to kind of cement them in your mind.

Anyway, like I said, I hope you don't take these comments to mean anything other than I just want to keep having these discussions and learning. Thanks again for taking the time to watch my replay, read my post, and offer your own insights.

zappazapper

@cheeseberry said in Game analysis - Replay #14886292:

Very intriguing read, love it!

I don't quite agree with some of your analysis but you already knew that it's not perfect, so who cares.

As long as you keep going at it with a mindset like this, you are bound to improve. So keep up the good work!

Well, I care!! I'd love to hear your thoughts on specific things you disagreed with, even if I still don't agree with you

zappazapper

@zokora said in Game analysis - Replay #14886292:

I would not invest too much time into those game analyses until you hit 1200+. Time would be better invested for the next game until you have some basic routines in your blood. Your mindset is good and you have a certain solid BO, you just need to get more game experience.

That's a fair point, but also, I'm a 41 year old man with a job in heavy construction and I often don't have the energy or mental stamina to play this game properly, while I certainly can watch a replay and organize my thoughts on it, and try and learn that way. Not only that, I think analyzing games is also a skill to be learned and if it's that important when I hit 1200 (not holding my breath), I want to already know how to do it.

ANALyzeNoob

@archsimkat
This is completely incorrect. You need to do analysis of your replays regardless of your current level, and I would argue it's especially important at low levels (0-1000), where there are likely large amounts of low hanging fruit, ripe for improvement. Getting "basic routines in your blood" and not analyzing your own gameplay while you are at 0 ladder rating will serve only to ingrain the bad habits you currently have.

I'm going to agree with Zokora on this one and say that especially considering how inconvenient it is to go through FAF replays compared to SC2, in particular being incapable of rewinding or skipping through to specific parts you are interested in, it's probably more valuable for a sub-1k player to just play more games and try to focus on important things in each game. You're obviously going to make mistakes, and vitally, should notice that you made IN GAME.

How does the noob even know what they are doing well and what they need to try to improve? IF they already DO know those things, they should easily notice them in game and not even need to watch the replay. "Holy cow I'm currently stalling -300 power in this game 5 minutes in, I should adjust my build order and make more power next time!" If you know stalling is bad, you realize the mistake IN GAME and watching it again in the replay is a waste. If you never pay attention to your resources, that's an even more important mistake that also ought to be realized IN GAME. "Hmmm, all of a sudden I have 10000 mass in the bank, last time I checked I was stalling -50. I should probably glance at my resources a lot more often!" I mean this is probably the most important thing to tell noobs. Besides learning a better build order, trying to manage eco better is something you should just know is vital, and don't need to watch a replay for.
E.g. It's better to cap t2 mexes then upgrade them to t3, than to just upgrade them to t3 without capping them. How does the noob realize this when watching the replay, without being able to notice it in game? If they don't know, watching the replay won't even reveal the mistake to them!
"Basic routines" are basically acquired from copying better players. I think that's an order of magnitude more beneficial to a noob than watching their own replays. Sure, you probably won't get too much benefit from trying to copy a 2k player's bo, but just look at someone a few hundred points above and that should be easily attainable.
The only reason you might benefit from watching your own replays is if you have almost zero awareness of what is even going on in the game, both economy-wise and with your units. You can become more aware from just learning the game better and getting more familiar with it, which I think happens a lot faster from actually doing it and seeing what better players do.

@Turinturambar
How do you want to check if you actually improved on the basics as you aim to improve on? Apart from finding new issues/new tactics in replays, watching replays can also just serve as a check how well your doing on your objectives e.g. how much e/mass did you actually stall/overflow and for how long? how many factories were idle and for how long? how fast did you actually expand? how good was your agression really? how good was your scouting/intel? how many tanks did you suicide for no reason?
While playing you will likely not notice how much you failed/not failed to fullfill your improvement goals.

You can check to see if you have improved by looking at your rating change over time. Your change after just a few games or a couple days isn't important, because there will be significant variance and noise in the distribution of results, but in the long run it's an excellent, objective indicator.
All of your other points get basically the same response as I gave to arch. Yeah, you might learn some new tactics from the other noob you played that beat you, but watching significantly better players than both of you is probably more valuable. If you're 500 watch the 800-1000s play and try to imitate that, or the best players that you can actually effectively copy (so as to not ingrain bad habits, which I do agree with arch could be a problem). And it can't be that useful to watch replays to see what mistakes you made, if you can't even notice you made those mistakes IN GAME. You won't do a good job of correcting your mistakes in the future if you still haven't learned where to focus your attention and realize where those mistakes are happening in game. E.g. "Oh wow I was stalling power bad, better change my bo." Ok if you notice that in the replay, you might improve your build order, but if you still haven't learned to very closely monitor your resources, when your opponent snipes all your power with an arty drop and you don't notice you have zero power income for 5 minutes, that's an exponentially larger problem that needs to be solved FIRST. And once you have figured out "pay close attention to your resources" you no longer need to look through your replays to see if you were stalling, because you will have learned one of the most important aspects of this game, which is always paying close attention to these crucial factors IN GAME. You see you were stalling, then ALREADY KNOW to make some adjustments for next game. Watching the replay is a waste. Just go have fun and play some more games. Who knows, if you miss that next ladder game by a minute, you might not get another one for 3 hours!

Edit: my examples were economy related but the same thing would go for units. "Hmm I sent all these tanks in to this unscouted area and they all died without killing anything. That was bad." Even if you didn't see the fight and just notice the army is gone, if you see point defense where you sent tanks, you realize you might want to micro those units or watch them closer and not suicide them. Even noobs can figure out "suiciding units=bad," but what they might not even understand fully in game, or from watching their own replays, is what types of strategies are most effective. I can imagine a noob thinking, "Oh, I guess I just wasn't at critical mass of tanks, and need to send even more next time!" Where they would instead learn from watching better players that attacking t1 pd with t1 tanks is not a strong tactic.

Turinturambar

why do you assume knowing an issue means you suddenly "magically" dont make that mistake anmore? The whole point in improving is picking 1-3 things in your gameplay you wanna fix and improve your habits on and then concentrate on them ingame for the next XXX games.
Watching your own replays from time to time is a good correction on your ingame perception - since ingame perception can be pretty bad - and serves as additional feedback. Ratingchange in this case is utterly useless since it contains anything and can even mean you improved in other areas than the one you aimed at improving.
In this regard watching replays also doesnt mean in detail analysis on -1 speed from both perspectives, but just e.g. to check stalls watch it at +2 or w/e from your own perspective and only look at ecobar for the entire replay and if there was a real issue slow down/rewatch to check why the issue happened (if even necessary). So you have an objective indicator on how much you actually improved on the thing youre working on.
Same can be true for intel/expansion/reclaim/unit movements etc.
When not improving on things systematically one by one you can run the risk of making all of them kinda mediocre which can still get you somewhat far but not beyond, so its in the long term less efficient.

ANALyzeNoob

@turinturambar said in Game analysis - Replay #14886292:

why do you assume knowing an issue means you suddenly "magically" dont make that mistake anmore?

Please read more carefully, because I never said that. " You're obviously going to make mistakes, and vitally, should notice that you made IN GAME." "You see you were stalling, then ALREADY KNOW to make some adjustments for next game. Watching the replay is a waste." I clearly never said noobs don't make those mistakes, just that they can easily notice them in game. Basically my point is noobs benefit more if they focus on learning general strategies, tactics and simple build orders and learn to execute them fairly well, but don't need to look at the replay to realize what mistakes they made.

I agree focusing on improving things systematically is probably more efficient to improve, but I'm saying watching replays isn't required for that, or the most efficient use of your time to do it. You hit the nail on the head: "picking 1-3 things in your gameplay you wanna fix and improve your habits on and then concentrate on them ingame for the next XXX games."

I would agree watching a replay occasionally is fine, especially when you can't find any games, but I'd probably only analyze less than 10%, probably about 5%. The largest benefit I see is to compare my overall performance to someone else to get an idea of who had a more efficient build order and possibly copy some of the things they were doing. But in that case I'm mostly watching someone else and not so much what I was even doing, just like I said before that they should just watch replays of better players and try to emulate them instead.

EDIT: So the TL:DR Conclusion is: basically if you can find a game, just play rather than watch a replay. If you can't find a game, there is no downside to analyzing your replays, but it is probably better to watch replays of better players instead of your own games.

maudlin27

I found my biggest improvement (moving from c.400 rank to 800) was primarily from watching replays where i got crushed and figuring out what my opponent did. Playing more games instead wouldn’t have taught me that anywhere near as quickly.

I also found sandboxing initial strategies numerous times to both figure out what works and to memorise it was better than just starting a game and trying to think about it on the spot (as I was frequently finding i would be close to overflowing mass despite trying to build more factories early) - sandboxing and replays helped me realise i was sending my acu to the front too early and that i should get it to build more factories first.

Jip

@CorvathraNoob You're essentially telling people to not analyze why they failed an exam, but just to make 10 more. In my opinion as a teacher, you're not learning something if you're not willing to take a step back and understand what you did previously.

arma473

Different people learn differently. There are things you will learn in-game that some other players won't catch unless they watch the replay. And things that you won't learn by playing the game. There are certain mistakes that you could make 100 games in a row without ever having a clue. Maybe for some concepts, the only way for you to learn it is if someone else just tells you. A one-size-fits-all approach doesn't work because people are different from each other. If you are benefiting from watching replays, then keep it up. If you think it's a waste of time, there are other things you can do.

Some chess players can recite their games from memory. They get out of a game and they could tell you a complete list of the moves both sides made. Other people can't do that. Someone who can't do that would benefit a lot more from watching a "replay" of their game, than someone who already accurately remembers everything that happened.

I definitely agree with the approach of picking 1-3 things to work on improving at a time. Zock himself said to pick just 1 thing, but that every time you go into a match, you always have 1 thing you want to test or improve.

And keep an open mind. It really doesn't hurt for ZappaZapper to think T3 land is silly and T1 spam is best. Sooner or later he will learn the value of moving up to T3 land. As long as he keeps an open mind, when his opponents eventually show him the value of T3 land, he can learn and adapt. In a sense, everything you learn from playing against 500-rated players is dubious. If you only watched games where FLOOD rolled 600-rated noobs, you would learn the wrong lessons about what works and what doesn't work.

If Blackheart or Thomas wants to take ZappaZapper under his wing and play 1v1s against him for 6 hours a day and explain the right way to do things and the right way to understand the game, that would be the best way to get Zappa up to 1900 as fast as possible. Anything else and he will develop bad habits on the way up the ladder. It's basically inevitable that people will get bad ideas and bad habits as they move up and part of moving up is not just adding things, but also discarding things.

zappazapper

@corvathranoob said in Game analysis - Replay #14886292:

EDIT: So the TL:DR Conclusion is: basically if you can find a game, just play rather than watch a replay. If you can't find a game, there is no downside to analyzing your replays, but it is probably better to watch replays of better players instead of your own games.

That is kind of the situation with me, as I often sit for over an hour waiting to get matched, and also like I said before, I often don't really have the energy to actually play because of work, and I'm not just trying to protect my meager rating, it's more that my goal right now is to improve and there's zero chance that I'll improve if I can't even keep my eyes open. But watching replays and writing an analysis can be done at a much slower pace, and just because I don't think I'm in the proper mindset to play doesn't mean I'm not still interested, so it's a better thing to do than nothing. And just so you know, I do watch other players replays, probably more than my own, it's just that I'm not writing analyses of their games.

biass

zappazapper said in Game analysis - Replay #14886292:

What's different about playing this game and playing a musical instrument is that playing this game for the first time is a little like picking up a guitar for the very first time and trying to learn "Flight of the Bumblebee". You really should just try and learn "Louie Louie" first or some other dumb schlockey song that you'll never want to be caught dead playing in front of humans (I would) because you're not ready to attempt "Flight of the Bumblebee", just like I'm not really ready to learn how to micro Inties at this point.

It's fine to think this way, but let me use another guitar analogy for you here. When you're learning to play a new song, you learn to play it at a very slow speed, but correctly. You don't play the song incorrectly in order to try and match the tempo right away. Once you commit the correct chord/fret/etc changes to your muscle memory, you can work through them faster and faster until you're doing it at speed. If you commit the incorrect notes/etc to memory. You'll definetly know that it's tough to work them back out of your playing when you are doing them subconciously.

If you do the right things now at a poor speed. It will be much more beneficial for you overall.
I'm not expecting you to go into a game and spam out 100 move orders to do perfect circle micro in your game. I only expect basic moves to shoot down threats and fighting enemy planes when you know you're ahead.

zappazapper said in Game analysis - Replay #14886292:

I can honestly say that reclaim is at the very top of the list of things I would like to improve on.

Reclaim is something you'll be working on thoughout your entire time playing this game. Right now you don't need to make it overly complicated. Ensure you get as much reclaim as possible by any means possible. Lots of people just end up taking engineers and doing attack moves to the other side of the map, for example..

zappazapper said in Game analysis - Replay #14886292:

"You shouldn't focus on your army movements or stuff like that so early on"
Mmm, that's a little bit of a contradiction there. If I don't need to focus on army movements at this point, why do I have to focus on micro'ing air?

You should practice it in your games, but you do not need to focus on it in your post game analysis.

zappazapper said in Game analysis - Replay #14886292:

A tank that didn't die because another type of unit gave me a slight advantage in one aspect over an opposing force of pure tanks is a tank that I don't have to build again.

Thinking this way is fine for other units, but a lot of the game and your choices within it often revolve around comparing your "X" to the "X" of the enemy. For example, you might count the number of land factories you have to see if you'll eventually have enough tanks to take map control.

While arty and aa are both:

Situational units that you might only build if X is present.
Still somewhat capable of contributing to battles
Cheaper, and take less time to build (somewhat)

Engineers are something else entirely. When your factories are essentially not making land units one 5th of the time, your ability to scout and make choices in the game become obfuscated.

I would suggest making factories exclusively for engineers if you need them. If you end up not needing the engineers later (it's hard to make "too many" engineers), you can always make more tanks instead. Or just reclaim them.

zappazapper said in Game analysis - Replay #14886292:

Anyway, like I said, I hope you don't take these comments to mean anything other than I just want to keep having these discussions and learning. Thanks again for taking the time to watch my replay, read my post, and offer your own insights.

No problem mate. I'm always around to answer questions

CheeseBerry

@zappazapper said in Game analysis - Replay #14886292:

@cheeseberry said in Game analysis - Replay #14886292:

Very intriguing read, love it!

I don't quite agree with some of your analysis but you already knew that it's not perfect, so who cares.

As long as you keep going at it with a mindset like this, you are bound to improve. So keep up the good work!

Well, I care!! I'd love to hear your thoughts on specific things you disagreed with, even if I still don't agree with you

Biass already talked about quite a few points that were on my list as well, the most important one being that Sup Com is indeed a game focused on macro, and not micro.

Or phrased differently: Supreme Commander is a game about getting more stuff.

First and foremost, this means that you need to spend your mass on stuff, because mass in the bank does literally nothing for you.

Just spending your mass is imo by far the most important thing to focus on.

Once your massbar is empty, how do you get more stuff? Well you need to get more mass. As a consequence: Supreme Commander is a game about getting more mass.

Why are mexes and reclaim good?
More mass and hence more stuff.

Why is map control good?
Map control leads to more mexes and more reclaim, which leads to more mass. (Also it gives scouting information, but nobody below 1k scouts, so who cares.)

Why are armies good?
Not only do they keep you alive, so that you can spend your mass on stuff, armies are also used to take stuff from your opponent and if your opponent has less stuff, you comparatively have more stuff.

Why are t2 mexes good?
If you don't sacrifice too much to get them, they will lead to more mass and hence more stuff in the future.

The only strategies that fall outside of the "get more stuff" plan are snipes, cheeses and memes. To be clear, all of them are valid strategies sometimes - many tournament games have been won by a well executed snipe after all - they are just not worth focusing on w/o having your fundamentals in order.

Note that I didn't even say what "stuff" you should actually build, because fundamentally it doesn't matter (yet).
If you have twice the total mass of your opponent and you just spend it on something that can attack, it becomes very easy to win the game.

In your analysis you are talking about unit movements, army compositions, attacks that did or didn't work, intie micro, tech levels, pgen adjacency, and so on, none of which matters (yet).

Your final bullet points almost hit the nail on the head as they are all different versions of "I failed to get mass when I had the opportunity."

Crucially though, you do have a full mass bar from minute 6 to minute 14 which you mention only once, even though it's the most important of all of them.

In short: This is a game about more stuff and by extension a game about getting and then spending more mass. Everything else is just details.

ANALyzeNoob

@jip said in Game analysis - Replay #14886292:

@CorvathraNoob You're essentially telling people to not analyze why they failed an exam, but just to make 10 more. In my opinion as a teacher, you're not learning something if you're not willing to take a step back and understand what you did previously.

Again, can people please read what I actually wrote? I never said people shouldn't analyze their mistakes. What I ACTUALLY SAID, hopefully for the last time, is that YOU SHOULD REALIZE YOUR MISTAKES IN GAME, AS YOU BECOME AWARE OF THEM. WATCHING THE REPLAY TO SEE IT AGAIN DOESN'T GIVE YOU NEW INFORMATION. If you aren't becoming aware of your mistakes in game, the most important issue is you still haven't learned where to focus your attention.

@maudlin27 "figuring out what my opponent did"
Exactly what I said: analyze what better players do.
Also, sandboxing build orders and analyzing replays are different things. I completely agree that sandboxing starting builds can be very helpful to remember build orders better and make fewer mistakes.

Jip

@corvathranoob

I did read what you wrote. And my metaphor fits. You want people to assess the game while playing. Similarly, it is telling a student to be able to assess a certain mathematical rule exists during an exam while he doesn't know about the rule, but lets get on with the next exam right after.

Inspecting your gameplay (or exam) and comparing it with your direct opponent (or answer sheet) can never hurt, whether you won or lost. It doesn't have to take forever - you can run it at +5 or +10 for all I care. Briefly taking a step back and understanding what you did wrong (or right) is better than just diving in right away again and again.

Talking with a trainer about your replay is like talking with your professor about your exam results .

Back on topic, I think it is great that @zappazapper is looking at his replays and trying to understand by essentially talking to his (graduate) peers right now whether his insights make sense.