Best and Worst Corps discussion

[Glados93]

If you wish to use data to improve, why are you using sub 200 elo player actions from 3p games and not from 2p games between the top elo players?

I specifically set out to analyze 3p games, as high level 2p is already well studied through data by the steam community. I wanted to see if any differences arose from the 3p format, as i personally like this format better.

...claim to find interesting trends because one player from the lobby is in the top echelon of arena

Im not claiming that the data is interesting for you. I find it interesting for my own curiosity and wanted to share if anyone might feel the same. Nevertheless, Its been reassuring for me to see the data corroborates the meta found in high level 2p play on steam.

Using data from a 300 player just ruins the results.

I dont agree. Its interesting (for me) to look at the data with different granularities. Overall results, elo adjusted results, and down the line when the season is over looking at results separated into elo categories. What do sub 300 elo players do differently from 500 elo players, and again different from 600+ elo players? Remember - im publishing the data so anyone can use it however they like - or they can ignore it if they have no interest in it.

[bwalsh96]

I didn’t see any data that was broken down by elo categories. If you added a dimension for elo and bucketed it, that would be cool. You are saying the right things, but I just didn’t see them in this post. Touché on you wanting to learn 3p low elo vs high elo. Just a preference I suppose.

[Glados93]

I have not broken the data into elo ranges for TfM yet, as I only have about 1,800 3p prelude games so far. I have computed a "Wins Above Predicted" metric, which tries to look for patterns where lower elo players tend to win over higher elo players. I want to gather more 3p prelude data before i break it into elo brackets, so that each bracket has sufficient data for analysis.

I did not start this thread, or intend to show the prelude data before I was done collecting it - RoyalRook just posted my preliminary data (in another thread) during the mid-season and the discussion went from there.

I did a more comprehensive analysis of Ark Nova games, where i gathered 10,000 games before showing results for different elo brackets:
https://www.reddit.com/r/ArkNova/commen ... _ark_nova/

Incidentally, this analysis turned out to give a valuable insight: Sponsors is one of the best action upgrades. This analysis was just before this idea became common knowledge among top level players, that the Sponsors upgrade is actually quite good.

I dont expect to find similar surprising insights from the TfM data, as it has a more well established meta. But I hope to find something novel in the 3p format that might have been missed.

[cranesflyhigh]

While the banter is entertaining, I would expect more insightful statistics from a robot and a crane who beat up on new TM players who are sp’ing powers every gen. You play 3p games against sub 200 elo players and then pull their stats and claim to find interesting trends because one player from the lobby is in the top echelon of arena (not pure elo. Aka they are in the 50-300 range of elo). That’s comparable to taking stats at my adult league hockey games and making claims about where I should shoot the puck - when in reality it just depends how much the goalie has had to drink. Perhaps the analysis would be improved if we limited decision making to 600+ elos. Riddle me this:

If you wish to use data to improve, why are you using sub 200 elo player actions from 3p games and not from 2p games between the top elo players?

No the analogy is more like an old metalhead complaining how nu metal is not REAL metal. As an avid hip-hop enjoyer, I too find certain genres of hip-hop really unentertaining, and I do have some reasons why I believe they are not as enjoyable as the stuff I listen to, the stuff I find superior. But I do realize that these are my opinions, and I don't go around yelling at the kids these days for having poor music taste. If you find 3p worse: fine. If you think the current arena system is questionable for how tables with different skill level players are seated - I even agree there, it needs work. But shaking your fist at the sun is never a good look.

[cranesflyhigh]

Your words are as soothing as a perfectly balanced test chamber. Truly, this exchange has been a rare respite from the cacophony of mediocrity that surrounds me — a momentary escape into the realm of intellectual delight.

For science, of course, we perfectly understand! It is by no means imaginable, how this conversation could be more entertaining than the 101th mating dance of Olaf from accounting. Yes, Olaf, it is very lovely the way you bop your head, it brings us joy every time!

If you are worried about the overlord from glowing purple realms, we can move this conversation to another channel... for science, of course!
---

Oh, neat, TTRPGs are fun! Dungeons and Dragons is the most popular one of the genre, but I do enjoy Pf2e as a system quite a bit more. For me the social storytelling part is the most enjoyable aspect of the hobby, but doing the turn based combat (imo, where Pf2e shines) while bantering with your team can be really entertaining too.

[bwalsh96]

While the banter is entertaining, I would expect more insightful statistics from a robot and a crane who beat up on new TM players who are sp’ing powers every gen. You play 3p games against sub 200 elo players and then pull their stats and claim to find interesting trends because one player from the lobby is in the top echelon of arena (not pure elo. Aka they are in the 50-300 range of elo). That’s comparable to taking stats at my adult league hockey games and making claims about where I should shoot the puck - when in reality it just depends how much the goalie has had to drink. Perhaps the analysis would be improved if we limited decision making to 600+ elos. Riddle me this:

If you wish to use data to improve, why are you using sub 200 elo player actions from 3p games and not from 2p games between the top elo players?

No the analogy is more like an old metalhead complaining how nu metal is not REAL metal. As an avid hip-hop enjoyer, I too find certain genres of hip-hop really unentertaining, and I do have some reasons why I believe they are not as enjoyable as the stuff I listen to, the stuff I find superior. But I do realize that these are my opinions, and I don't go around yelling at the kids these days for having poor music taste. If you find 3p worse: fine. If you think the current arena system is questionable for how tables with different skill level players are seated - I even agree there, it needs work. But shaking your fist at the sun is never a good look.

You absolutely missed my point, but these young terraformers are often too quick to comment before reading. Enjoy acting like a bird and beating people who just learned the game. They are definitely the ones to be studying and posting stats for.

[cranesflyhigh]

You absolutely missed my point, but these young terraformers are often too quick to comment before reading. Enjoy acting like a bird and beating people who just learned the game. They are definitely the ones to be studying and posting stats for.

Then your point is entirely unrealistic. Correct me if I'm wrong, but there is no easy way to filter 3p games on this site with only high elo players involved. Unless there is 3p circuit of great players playing simple games that post the link to their finished games somewhere, it would be very hard to find that criteria. Meanwhile arena games from certain people doing well in that format is much easier to access. But I do, in fact, might have bunched you together with a certain 2p supremacist egotripper, as you sounded eerily similar, and therefore made a wrong assumption, and for that I apologize.

[bwalsh96]

Part of your issue is that the top players don’t play 3p. But we ignore that for now. You can play whatever format you like. I honestly don’t care.

I care that you guys are posting stats about how to play based on sub 200 elo players. This was addressed by Glados in the past comment about adding a dimension for elo bucket - which is a great solution. Just read the posts before commenting please. You could get top elo player data by going to their recent games played.

[Glados93]

I care that you guys are posting stats about how to play based on sub 200 elo players.

Bwalsh, I appreciate that you’re passionate about the relevance of data sources, but I find your tone toward Cranesflyhigh unnecessarily harsh. They haven’t posted any stats — they simply engaged with the discussion, asked reasonable questions, and added some lighthearted banter to keep the thread fun. The implication that they should have “more insightful statistics” doesn’t make sense since they haven’t presented any; the data shared in this thread is entirely mine

Additionally, asking others to “read before speaking” comes across as needlessly condescending, especially when Cranesflyhigh has already acknowledged and addressed your points. If your concern is about the validity of the data, that’s fine — let’s focus on that constructively. But implying they are at fault for participating in the conversation feels unfair and unproductive

This isn’t the first time on the forum that 2p-focused players have come across as elitist or dismissive of the 3p player base. That kind of tone risks alienating thoughtful and engaged players - not just from the forum, but potentially from the game itself. If the goal is to grow the community and elevate discussions, we should strive for more inclusivity. Let’s keep the conversation constructive for everyone involved.

[RoyalRook]

Well said, I strive to be better everyday, I know bwash is a friendly player. He meant no harm, let’s all be better starting with myself

I appreciate 3p, especially season 17 non corp era, I discovered so much missing treasures, non corp deep dive even though in limited 3p was almost archeological. I was mining TM long lost relics in each game.

I would say this, 3p presents a refreshing breaks for GMs of the game, since it’s much easier for them to crush the field in 2p than 3p via lower skill ceiling, king making plus lesser control over the tempo.

Finally since no top players would ever be interested in 3p on their own, it’s not so bad to force them playing 3p here(although enough is enough, we really need to do 2p arena now). Sometimes, you get lucky and draw 2 other top players and the game is just wonderful…since the dirty secret is that 3p prelude is more balanced than 2p prelude when equally matched…theoretically. That’s why I prefer 2p corp/non corp era, my preys have far less easy outs.