Author here: I really appreciate the nuanced and constructive discussion in these comments -- this is the reason I post to HN. Thanks!
> This is sad that the article goes on to explaining the crucial importance of smokes, flashbangs, etc while not focusing on them for analyzing games.
There are several layers here and I should write a follow-up article to discuss that because I realise now I was stuck in my mental context too much to explain the bigger picture in the article.
This article is not, strictly speaking, trying to explain performance; it tries to explain the Leetify rating which is hopefully a fair proxy for performance. Nor is it really about giving details on how to achieve a high rating -- that requires the things you mention to put oneself in position for a favourable engagement. (In fact, one can think of the kill itself as a proxy for several other things, such as successful flashbang usage.)
This article focuses entirely on one specific kill-or-get-killed decision. It took that much analysis to provide a useful guide for that decision alone! A more complicated set of decisions I don't even know how to approach quantitatively.
Like, how would you define a successful smoke, if you had to do it based on anything that does not boil down to community feels?
I'll look into scope.gg (and tracker.gg mentioned elsewhere in this discussion) and see if I can be inspired to analyse other decisions. Thanks again!
Thanks for the clarification, I see that I misunderstood your article as "I’m dissatisfied with existing tools so I’ll make my own", while it’s more along the lines "let’s reverse-engineer leetify". And for the latter it’s an excellent one :) Thanks!
Indeed k/d is an overall good proxy when taken with a step back, eg for a whole game. When looking at a single round, it ends up being noisy. Which is not to say it’s useless! In the end this is a shooter game, and to shoot is in the name (and not to die but that’s another matter) :D Now to the matter to evaluating odds of winning a combat.
To develop a bit on smokes: with some domain-specific knowledge you could approximate its usefulness. Smoke in catwalk on D2 allowing T to go through mid while a CT stands on shirt unable to take info nor shoot? Now that’s a good smoke. But this would require very tedious work an be non exhaustive. You could make a rating of "how many people walked in front of an opponent thanks to a smoke". Or how many bullets missed a CT defusing in the smoke. Possibilities are endless but not perfect.
For example scope.gg rates flashbangs by how many people were blinded. And I’d not be surprised if they gave a negative ranking to flashbangs that were heard but did not blind anyone: those give info to the opponent without helping you.
Now if you made a follow up that’d be great but you’ll end up meta-gaming instead of gaming. But that’s fun too!
> For example scope.gg rates flashbangs by how many people were blinded. And I’d not be surprised if they gave a negative ranking to flashbangs that were heard but did not blind anyone: those give info to the opponent without helping you.
Not sure if you'll read this, but I've been trying to figure out how the scope.gg "Rating 2.0" is computed, and all they say is that it's similar to the HLTV 2.0 rating, which in turn basically decomposes into kills and deaths and proxies thereof, i.e. no contribution from e.g. flashbangs. This would be in contrast to Leetify which at least counts flashbang assists.
I haven't used it in over a year so my memory may be failing me.
And to be fair I used their match explainer that focuses on rating in a single round, and to show you your mistakes. They rate pretty much everything, including the flashbangs I described. I must admit, I was not interested in my ranking, only in my mistakes. It is very possible I assumed all ratings in a match analysis counted towards your overall rate, but appears to be wrong.
So yeah, in the end we are not talking about the same thing. Sorry for bothering you with this!
(although it would be pretty interesting to have all those fine details be at least a signal for a player's overall rank)
> This is sad that the article goes on to explaining the crucial importance of smokes, flashbangs, etc while not focusing on them for analyzing games.
There are several layers here and I should write a follow-up article to discuss that because I realise now I was stuck in my mental context too much to explain the bigger picture in the article.
This article is not, strictly speaking, trying to explain performance; it tries to explain the Leetify rating which is hopefully a fair proxy for performance. Nor is it really about giving details on how to achieve a high rating -- that requires the things you mention to put oneself in position for a favourable engagement. (In fact, one can think of the kill itself as a proxy for several other things, such as successful flashbang usage.)
This article focuses entirely on one specific kill-or-get-killed decision. It took that much analysis to provide a useful guide for that decision alone! A more complicated set of decisions I don't even know how to approach quantitatively.
Like, how would you define a successful smoke, if you had to do it based on anything that does not boil down to community feels?
I'll look into scope.gg (and tracker.gg mentioned elsewhere in this discussion) and see if I can be inspired to analyse other decisions. Thanks again!