Kinch Analytics TM

🏎️ Trackmania Analysis 🏎️
📊 Stats and Insights 📊

[Kinch Analytics TM @KinchTM]

My stats rate this round from yesterday as the best this season in terms of times driven.
The 4 times had expected placements of 1.02, 1.05, 1.09 and 1.38
https://t.co/wIp3oNOZQv

[Kinch Analytics TM @KinchTM]

@Hisoris I suspect that if you look at teams that lose both 2nd picks, or both 3rd picks, you would see a similar match loss rate.
To show that 1st pick is the most important you would likely need a much bigger sample size, and then compare to what I just suggested on 2nd/3rd picks.
[2/2]

[Kinch Analytics TM @KinchTM]

@Hisoris I think you've only shown that winning maps is important, not that 1st pick is more important or anything.
You need to win 4-1 from there, so if every subsequent map is 50/50, you have 81.25% loss chance, which matches well enough with the results [1/2]

[Kinch Analytics TM @KinchTM]

TMGL Map Win Rate by Scoreline

2-8 Comeback: https://t.co/QDQ9d3cKYT
1-7 Comeback: https://t.co/Hro0Zd3MmC

[Kinch Analytics TM @KinchTM]

🏆 TMGL - Stage 1🏆
💯 Player Ratings 💯

🥇 @PacTM2
🥈 @Affi1606
🥉 @CarlJrtm

Explanation and comments here:
https://t.co/FzEytseiSU

[Kinch Analytics TM @KinchTM]

📊 TMGL Clutching and Closing📊

Everything is explained here: https://t.co/EwaIBPnrIf

[Kinch Analytics TM @KinchTM]

@skins_tm Good question. One I shall have to answer another day!

[Kinch Analytics TM @KinchTM]

The points required for Top X on Campaign has hardly changed despite a huge increase in the number of players.

[Kinch Analytics TM @KinchTM]

@MXFlamezTM Reassuring, but then you realise that you actually have to keep improving to get the same points thanks to each map having more and more good players on it each season (something that I have yet to prove, but is surely true)

[Kinch Analytics TM @KinchTM]

I actually expected these to decrease over time because being consistently Top X across all maps should be much harder now than it was, but I guess that the 3-month window allows the best to always achieve great things (if they can be bothered with campaign haha).

[Kinch Analytics TM @KinchTM]

My calculations rate this as the best TMGL round so far in terms of the times driven
It is Playday 4 - G1 v ITB - Agility Dash - Round 1

4th place here would have won over half of all other races on the map.

https://t.co/nr59XT6xYx

[Kinch Analytics TM @KinchTM]

TMGL Players definitely show a bit of a warm-up effect as a map progresses, with the later rounds rating better than the earlier ones.

Round 6 appears to be the sweet spot!

[Kinch Analytics TM @KinchTM]

👨‍❤️‍👨 COTD Rivalries 👨‍❤️‍👨

@ScrapieTV might have a big lead over @GranaDyy in total COTD wins, but GranaDy has a slim lead in their head-to-heads with a 7-6 record.

The biggest "whitewashes" are 5 pairs that have a 4-0 record

[Kinch Analytics TM @KinchTM]

TMGL Round Outcomes
🎯Aces: 132
☑️ Wins: 169
↔️ Draws: 136

If you're interested to see my analysis of why Wins is more common, I wrote a long thing about it here:
https://t.co/VtORUUi6u0

[Kinch Analytics TM @KinchTM]

🌍 COTD Countries since 2022 🌎
🥇 Germany
🥈 France
🥉 Belgium

Not too different to the previous tweet, but 🇫🇷 were a little down on their 2020/2021 stats and 🇬🇧 is moving up

[Kinch Analytics TM @KinchTM]

🌍 COTD Countries 🌎
🥇 Germany
🥈 France
🥉 Belgium

- The big 2 have won over half of the COTDs ever
- 2022 stats coming tomorrow...

[Kinch Analytics TM @KinchTM]

My Adjusted Average Rank:
- Compares times across all matches, adjusting the actual race rank to fairly reflect how fast/slow the drive was.
- Weights performance slightly more heavily in the more important situations where players need to clutch or close out a match.

[Kinch Analytics TM @KinchTM]

📊 TMGL Player Rating 📊
🥇 @Affi1606
🥈 @PacTM2
🥉 @gwentm4

My Adjusted Average Rank metric is explained in the replies

[Kinch Analytics TM @KinchTM]

Basic average race rank has a few problems:
- If your teammate is strong/weak you have a much harder/easier job
- Each team has played different teams so some have had an easier schedule so far
- Not rewarded for winning by a long way and punished a lot for losing by thousandths

[Kinch Analytics TM @KinchTM]

🔟 COTD Top 10s 🔟
🥇 @ScrapieTV
🥈 @GranaDyy
🥉 @Marius89TM

- Scrapie wins 28% of the time (88/315) when he reaches Top 10, while GranaDy has a 23% conversion rate
- Marius confirmed as a safemeister, but Down is right up there too