You could probably get 95% of the way there by pulling down the paid roster for each team, maybe?

I didn't think about that, I'll try it out next chance I get

Sheet updated to include week 3 matches, I hope everyone had fun playing koth this week c:

ty for the effort put into this. I'm wondering if some of the minimum DPM scores are caused by ringing? Sometimes after you leave logs displays a 30 min playtime next to your name.

AM div had a few matches overturned today due to a player alting: Keaghan. What I heard the original account probably was: XBS.

There was a match in week 2 that had this issue, I simply did not record that player's stats seeing as they played for 30 seconds and left. It's entirely possible, however, that I missed more of those cases. Shouldn't be too hard to fix if we find them.

If the medic scores are calculated across all games, using total deaths really breaks down because not playing is automatically better than playing. Ringers who only played one game are by far the best by default.

Just use stats/30m so it’s normalized

btw why is the rank 1 scout a ringer

Cuz a single good map vs a weak team will give you great stats, and when your opponents are normalized across a large number of matchups you look better. Ringers generally have either really high or really low stats compared to consistent players.

i didnt mean like how is he ranked 1st, im just wondering why ringers are even on the spreadsheet in the first place. if its a power rankings for the div, then including someone who isnt in the div and is only going to play like once doesnt make sense.

also i didnt realize but 1st ranked med is also a ringer

Sorry, been a bit since I checked the thread.

Based on how we gather match data, ringers have their stats pulled from each log automatically. Although identifying ringers isn't necessarily difficult, modifying our score function to exclude their data is. If any spreadsheet wizards know of an easy implementation, my dms are open (melons_).

We have now seen the flaws that come with using total deaths as a metric. The original plan was to compare stats week by week. Now that we've moved towards analyzing each player's data overall, combining total deaths is a clear detriment to those who've played more matches. As such, we have made the following changes:

• Medic deaths are now averaged over their total matches played
  
• Minor adjustments to score calculations when players play more than one class (i.e. playing medic one match, scout the next).
  
• Theoretical Super Team calculation is now fully automated
  
• Added "Runner-Up" players for each class

Good luck with matches this week, everyone! I love reading all your suggestions feedback, so don't quit c:

Dividing by matches played is worse than a set time cuz match lengths are variable, so playing a 50 min banny config game is gonna result in more deaths but not more deaths/30 min, so that might be a better metric

Alright here's another quick one:

• Match durations are now used to normalize a player's Deaths/30m. This is currently used in calculating Medic scores.
  
• Damage Taken/Minute is now used to calculate Scout/Soldier scores (using a ratio of DPM/(DT/M)). The intention with this change is to equalize Combo/Flank scores.

I had to redo the pulls for each week's matches (week 4 soon tm), so I really hope these changes are ideal. As always, let me know if there's any other optimizations you wish to see, my dm's are open as well o7

Sheet updated to include Week 4 matches o7

Thanks for keeping this up, it's injected a little fun energy into the season.

But please, please replace Diop with Jaguar on scout our team so I stop hearing half my team complain about it in scrims every night.

Truly sorry about this! When I rewrote the sheet from scratch, I must have kept diop around despite replacing him after the first week's iteration. If anyone else has roster changes that need to be made, dm me on discord @melons_!

Sheet updated to include Week 5. o7

Can you replace Herb with Neo on the Seagulls?

A little late, but the sheet is now updated to include Week 6 B]

5/6 Theoretical super team members are now not main roster players!

Week 7 Update! Shoutouts to gleeby deeby vs i got nothing... crazy match!

With the regular season coming to a close, I've put out the final updates to the spreadsheet. I plan to make separate tabs for playoffs, but for now I'd like to thank everyone for all the feedback you've been giving; we had so much fun sharing this project with all of you!

As always, please let me know if I've gotten your roster wrong! @melons_

just want to throw a quick plug here: i am trying my best to cover IM playoffs on my twitch channel. Tonight we are going to be casting the Griselda vs Scringus Goes to Rehab game at 1030pm.

In addition, the tf2 team vs DAD 2: Back with Milk, and the Monolith vs Instagib matches are available as VODs.

See you guys tonight.

edit: nvm fireside are doing the griselda vs scringus game lol

reminder to get your demos in for the frag video!

tonight Loot and I will be casting tf2 team vs Scringus Goes To Rehab @1030 on my twitch.

Unfortunately I wasn't able to find someone who could do the other game on their channel (if you are interested DM me on discord @ just_jazz_ and I'll help get you a setup). I'll do my best to provide live updates of their match scores as the night wears on.

Good luck to all teams tonight!

Gunden is casting the other game tonight. Thank you Gunden!

https://www.twitch.tv/hone_the_rat hes back

Nearing the end of the season now, is it normal for AM to have this big of a skill gap? Some teams feel like they haven't touched tf2 before while others seem to have a pretty decent grasp of things.

Just wondering what others think - GL with everybodies final matches

yes