New stats page

[h3rmit]

As many of you may already know, for the past few weeks I've been running a new, standalone stats page for ENT stats in parallel with the official one:

https://h3rmit.gr/stats_ent/

The page synchronizes with ENT stats every 5 minutes. You'll see your games there with a few minutes delay. It is still in active development and a few corrections and new features and may be added over time.

You can see at a glance a player's performance per game, per hero, per month.

Performance is ranked in a scale from "Bad" to "Great" by comparing farm, kills, deaths and assists to the long-term community averages for the same hero, scaled to the game's duration.

For those who have already checked it, you may want to have another look at your stats after today. An important update was made to address an incorrect calculation of farm rating, causing a slight inflation in all ratings. Now everything is back on scale with the community long-term averages per hero.

A few notes:

• A few heroes (like Kaolin, the Earth Spirit) may be missing their icons and fail to get a correct rating. This happens in all ENT stats pages. It will be corrected at some point.
• A few items (like Aghanim's scepter on Treant) may also be missing.
• Older games (more than a year ago) may be missing their items, or hero level.
• Very old games (prior to 2021) may appear in your playtime and game count, but there are no stats for them.
• Links to players with certain special characters in their name may not work, and instead redirect to the search page. This will be fixed soon.

I'd like to hear your feedback, thoughts, suggestions and bug reports.

[dumpster_]

Is there some kind of adjustment based on how balanced the teams are?
Even the greatest players can sometimes perform atrociously, not necessarily through any fault of their own.
Even if one were to play perfectly it's still possible to end up 0-5.
I've seen it happen to every top tier player without exception.


It seems to me like there's enough data to do some machine learning, for example to predict the winner based on the players and hero selection.
I think it's pretty likely that you could achieve an accuracy of over 90%. (edit: 90 is too ambitious I guess, but 75% seems reasonable)
If one is successful with that, then the next step could be to evaluate how much a given player's presence in a game affects its expected outcome.
If you're really ambitious, you could even build an entire new rating system based on that.
Doesn't even have to be hooked into ENT in any way; it can just exist on its own for entertainment purposes.
Maybe nublets like venshitta will finally realize that the reason all their accounts inevitable end up below 1000 is not because of shitty teammates (which we ALL get) but because they're objectively dogshit. Hi boots first necro


A big obstacle towards both of those goals is the prevalence of smurfs, but I think it shouldn't be too hard to accurately determine who's who with some machine learning.
Of course one could trick it if they really want to, e.g. by changing their hero picks, the time they play, and the way they talk.
But why would anyone bother?
It shouldn't be too hard to reasonably accurately determine who's who after even a dozen games.
After >50 games I think you could achieve close to 100% accuracy.
You could also ask the admins to share the aliases with you in private to make training the model so much easier.


I've been thinking about this for years but the chance I actually do any of it is close to nill.
Do you have any grand plans like this?

[GodSaveTheQQs]

Very good analysis on the subject Dumpster_ and i totally agree with your thoughts if Hermit could do a system as such.

It seems to me like there's enough data to do some machine learning, for example to predict the winner based on the players and hero selection.
I think it's pretty likely that you could achieve an accuracy of over 90%.
If one is successful with that, then the next step could be to evaluate how much a given player's presence in a game affects its expected outcome.
If you're really ambitious, you could even build an entire new rating system based on that.
Doesn't even have to be hooked into ENT in any way; it can just exist on its own for entertainment purposes.

Thats why in a previous post regarding the system info Hermit provided, i made a thought about a new system based on "general efficiency" mostly. How much your playstyle, your moves in and out of a battle affects the whole game and the outcoume of the winner team. Dota is a game of mistakes after all. If Hermit could make a system rating based on the above plus some other factors as well, would be super cool i think!

P.S. H3rmit now you changed the system again i dropped from an average - above average player to average - below average even i have only one acc wich is 6 old

[Frank]

dont worry godsave im worse than you

https://h3rmit.gr/stats_ent/player/yamal

[GodSaveTheQQs]

Yamal is actually a wonderkid. So those stats are wrong brother. Maybe its a bug H3rmit should fix.