100 days of ladder
@tagada said in 100 days of ladder:
If you would be interested I could give you a short training to give you some pointers, we could then repeat that every say 2 weeks and post the training sessions on the FAF's YT and make it into a mini-series. I think it could motivate and help some people that are considering getting into the competitive side of the game by making sure that they know what to focus on and how to improve quickly.
For replays you can just share the replay ID, don't think there is a need for a zipped file.
@Tagada i would be very appreciative if you were to make this into a series im struggling to find a high rated player to learn off of
I like this series a lot. The self reflection is key to improvement IMO. You seem to be catching some of your mistakes and are working on fixing them which is awesome. Wishing you luck as you continue.
If I may offer some unsolicited advice:
I am not sure if this is a misinterpretation due to language use, but something I noticed is a few of your writeups you will say something like "I lost 'x' because my opponent did 'y', or 'I sent too many engis to location 'x' instead of doing 'y' with them. Where the implication is that 'x' was really bad and not to do it again. This is not always so black and white. The first 3-5 minutes of the game can be really dicey in terms of opening builds on certain maps, and fog of war is a thing even in the later stages. You can't always act on perfect intel and calculated risks are a thing from your opening build to what you do at min 15.
I really don't think you are doing this because your writeups seem like you are giving your gameplay some good thought, but I just wanted to make sure that you are not jumping to conclusions with cause and effect If you made a decision that is good 80% of the time, I want to make sure you are not discounting it because it failed on that 20%, without trying to understand some more context behind things that go wrong at face value. Remember, a good choice will not always yield a good outcome.
The one main piece of advice I want to give you specific from me is for you to ask yourself when you play the following:
What am I doing?
Why am I doing it?
What is my win condition?
I don't have good APM, I have decent micro, I have passable macro, and my late game is garbage. But I know how to find my win condition and exploit it. I highly recommend you try and remember when you analyze (and play) your games to look at the big picture, and not get lost in the weeds with micro/silly mistakes. Find your win condition and commit to it. But try not to get blinded to what is around you as you do so. Its very general advice, but IMO its the most important thing you can do in this game to win. Don't just play on autopilot. Play for a goal, and make it the right one.
Thanks Tex, this is really good advice actually, a big problem I have been trying to work on is actually figuring out the reason why I lost (or sometimes even the reason why I won), and why a decision was bad, rather than if it was bad or not. As you mentioned, I need to extend this mindset down to the individual decisions I make ingame as well. I have also noticed in certain games that I don't commit to a strategy or win condition, this is another area that I can work on. Many of my wins have simply been my opponent suiciding, but that is starting to change as I face stronger players where I need to take the initiative to win.
Maybe one day if i can find the time I’ll try and link a vod to a replay review of yours if you are fine with it being posted here.
@Tex that would be highly appreciated.
Day 10 (May 9th): Some more success today, although admittedly against a lower rated opponent. Today's game (Replay 17002038) took place on a very mass heavy mapgen map, something I could use a lot more experience on. From the start I was trying to follow the advice about what my strategy/win condition. would be. I decided on running a mix of T2 while I quickly scaled my main factory to T3 (not exactly the most original plan I know). I would use this to secure the corners which is where most of the reclaim would be. In the end my opponent somewhat underbuilt on tanks, and didn't scale as quickly, allowing me to catch and cook his ACU. It took a while to outscale him on land as I prioritized building way more engis to reclaim than he did, but eventually it paid off with the win.
I think I played this better than I would have at the start of this journey. I got most of the reclaim on my side, although there were a few gaps. I made an effort to scale up much quicker, and I made an effort to move engis to new reclaim fields before I stalled mass, rather than after. I mostly stayed somewhat balanced, although I had a brief stall while making my first t2 pgen.
I also made the mistake of not assisting my main factory with engineers. Overall I made too little buildpower in my base. I think I tricked my mind into thinking the buildpower would be used in my base, all while my brain went on autopilot sending them to the rock fields.
Overall though, I wasn't really punished for the mistakes I made, and the game ended before I could see any differences in scale. I have been looking for a replay to specifically ask for a critique. I think this game might be it, as I am not used to games that scale up this quickly and I am not sure how the game would have gone from here. I want to see how my analysis of even a short and simple game is different from other players. Would anyone be willing to watch this game (replay 17002038) and tell me what I could improve?
@evan_ On 17002038 i would say u have a hard time spending your mass early on which is bad like make 1 land fac first engie go reclaim and make like 8 pgens with acu and engies to have enough e to spend all the mass
Day 11 (May 12): Played 2 games today, and lost both. Both games seemed to highlight difficulty in defending against raids, even when I had tanks defending.
In game 1 (replay 17014918), a combination of an early powerstall, losing fights, and and investing in the wrong things cost me the game. I have figured out that I tend to overflow mass or hold large amounts in storage a while after a mass stall, this is often due to overcorrecting. I think I need to do less overcorrecting, even if it means mass stalling for slightly longer by not canceling and pausing every project. In the end, I had to keep my ACU involved in combat instead of upgrading, and investing in t2 mexes and t2 land put me far behind in t1 land, resulting in me getting overrun.
Game 2 (replay 17015030) was on Emerald craters, which is quite a tough map to play. Emerald craters is quite spread out, making it difficult to defend expanding engineers. In this game I did not attack as much as I should have, and was not able to defend against raids from bombers and labs. This meant a heavy loss of engineers which snowballed into being outscaled. I also underestimated how important land would be in the game. I think I believed land would not be as important as air. I was able to keep up for a while and take back and hold half the map for a while, however I was never able to make lastind damage on my opponent, and was eventually lost on all fronts.
Today's review is fairly short as I am somewhat tired. I have asked Tagada to train me tomorow, and will update you guys on the results.
Day 12 (May 14): Well it has been a couple days now hasn't it? A few days ago I trained with Tagada, and he showed me some problems I had been overlooking (as well as some problems I hadn't been working to fix hard enough).
It is hard to summarize everything in a short post, and I am sure @Tagada can weigh in if I missed anything, but the few of the main points are as follows:
-
My rally points are inneficient, often leading units to the wrong parts of the map or even suiciding them into enemy tanks. Tagada mentioned that rally points should typically be just behind the front lines (where your map control ends). This seems obvious in hindsight but I suppose I had been manually correcting where each unit goes all along, wasting precious time and apm.
-
My unit queues have also been suboptimal, as I have found myself not removing labs and t1 transports. In most of my factories I have been building a mix of tanks, engineers, etc which can be somewhat difficult to manage. Tagada suggested having some factories for only tanks and scouts, and some producing engineers. I have tried this, and found it quite a bit more convenient. He suggested having factories assist other factories to copy the unit mixes and rally points. This saves time and allows me to change the mix and rally points for all of them at once.
-
I react very slowly to things: Quite often when watching replays I see myself stalling for minutes on end without correcting it, or allowing my opponent to take out my expanding engies for free.. There isn't any magic solution to this other than to try and plan out things to be more effective ahead of time. I have been trying to have more game awareness for a while and will update you guys on the results.
-
Eco management and decision making needs some work. Often I find my mistakes and bad decisions compounding on each other. For example, making less build power than I need will cause me to spend less mass than I should, building up a large bar of unneeded resources. With nothing else to put it into I absentmindedly put the mass into mex upgrades when I need tanks. Another basic thing Tagada had to explain to me was how to tech transition. T2 units are generally more efficient than t1 and so once you upgrade you generally want to scale back t1. In my games I found myself upgrading to T2 land while also building new t1 factories to spam tanks. This is a somewhat unfocused and conflicted plan. Tagada explained that I should either push for T2 and not make any new t1 spam factories or I should stay on T1 for slightly longer and try to gain an advantage there before heading to T2.
There are many other points he touched on, which will hopefully be displayed in video format soon. As for today's games, I tried to apply the lessons, however they were somewhat anticlimactic. In game 1, (replay 17038956), my opponent raided me and then after a brief fight overextended with his ACU. I made more of an attempt to catch stalls/overflows (I'm also starting to learn when I stall the most and why). I tried to make more of an effort to protect my engis although I still lost 2 of them to a single selen, which is not exactly optimal, but oh well.
In game 2 (replay 17040925) I pretty much doomed myself from the start. It was on a large 20x20 water mapgen. I went for a first transport which was promptly scouted, locked, and shot down. I got 4 engis off and quickly laid down a factory, however, building a mex first combined with a bad powerstall prevented me from completing it in time, and although I tried to spread out my engis, evidently it was not enough as I stll lost them all in one pass, right before the factory was completed. Combined with losing air the game was lost as long as my opponent didn't make a major blunder. I somehow took the main central island back and won a ton of reclaim, but even with that there was pretty much nothing I could do, and in the end I ctrl-k'ed.
I've decided to dedicate some time to practicing and learning some basic build-orders, as well as doing some sandboxing to figure out where and when I make decisions that lead to stalls or overflows. Fortunately there is tons of new interesting content from various other players, so I have a lot of resources to help me. I'll update you guys on how it goes.
Hi guys, sorry about the increasing gaps between posts. This was a big commitment but I still intend to live up to it. I've put laddering mostly on hold while I search for a job. I'll try to play and post when I can so I don't lose my progress.
Hi Evan, still great to see this series even with some gaps. Really good way to improve, and it's inspiring me to go pick up some ladder to do the same thing. Laddering does seem to give more direct feedback on your gameplay, because any mistakes you may make aren't buffered by team mates, and vice versa. Looking forward to read more about your progress in the future.