First of all it is really hard to use statistics to properly measure anything if you do not have enough correlated data.With that in mind lets dive into the article which lead me to post this and the reason why their statistics are misleading:
https://www.invenglobal.com/articles/13583/lcs-spring-split-beyond-the-numbers-tsm-and-powerofevil-are-a-match-made-in-heaven (I do not think that this article is generally bad, but I disagree with their interpretation of their statistics)
The post states: "If there’s any evidence that POE outperforms Bjergsen, there should be pieces of it somewhere below:" (There are also some more statistics which are similar)
|Career Statistics||Azir WR||Azir KDA||Syndra WR||Syndra KDA||Viktor WR||Viktor KDA|
The then later claim "If we’re to look at only numbers, POE is actually a straight-up downgrade to his iconic predecessor"
So lets assume their claim is true what would the data in the above table need to provide us with the assurance that the only difference when gathering the data was the player and nothing else. This implies:
- League of legends is a single player game or decided only by the mid laner (as otherwise we would need to factor in all other players)
- Champion winrates are independend of matchups or all matchups when collecting the data were exactly the same.
- Pick ban phase has no importance on the result of the game and the data about the player
- There is no shotcalling or all shotcalling was done by the midlaner
If any of the above implication does not hold true you need to gather more data to get a proper result.
It can be seen at this point that the comparison is lacking so what could be done to check who is the better pro player?First you could try to incorporate the overall win rate of their teams and maybe their performance with and without that player (at best without any other roster changes)
If somebody knows where I can gather all kinds of stats about pro players in their career I could try to do a better comparison (Still not a good one as I think data is lacking and it is most likely hard to find good metrices).Also feel free to drop what you would consider good methods to compare pro players in the comments
Edit: To the people saying that statistics tell something, that I am just as bad as I have taken things out of context or that I have not read the article and am actually telling the same:
- I have read the article fully before posting which is indicated by my second quote which is from almost the end of the article
- What exactly do I disagree with?
For that first the structure of the article:
A question which needs to be answered: " Why did Bjergsen sign such high, very definitive praise of PowerOfEvil? "
A look into statistics which first compares the whole career stats of POE and Bjergsen with the conclusion: "While our short historical trip hints that Bjergsen has been overall better on control mages than POE, it led us to a new theory and a (perhaps shocking) revelation — that despite his inferior statistics, POE fits current-day TSM more than Bjergsen the player ever could. "
Third a section about other reasons (apart from the before mentioned statistics) why POE fits the TSM team.
Lastly a conclusion:
" POE is not Bjergsen. If we’re to look at only numbers, POE is actually a straight-up downgrade to his iconic predecessor AND to his days on FLY and CLG, where he looked way more threatening individually. But at the end of the day, POE has clicked with the TSM roster — a team that wants to play around these scaling control mages — at a level Bjergsen was unable to achieve on a consistent basis despite a superior level of skill to POE for the majority of his career.
The historical statistics might disagree with Bjergsen on the “Who’s better” question, but at the end of the day, POE doesn’t need to beat the numbers. TSM looked to find a Bjergsen replacement in POE and so far, he and his control mages have fit the new era of the team like a glove."
My point: While they also say that the end result cannot be explained by numbers my problem was the usage of those statistics to try to explain how they compare as the statistics don't factor in a lot of other important information. I am completely fine with the rest of the article.
- I do not think that numbers do not tell us anything it is just that the context is important. And if you present something like in the article not everybody will be able to tell how they have to interpret what they are seeing.
If you still think I am biased or interpreting the article incorrectly please drop some more comments.
TLDR: Do not try to compare apples with pears or if you do consider all the differences between both.
Source: Original link
© Post "How to use statistics incorrectly to paint a misleading picture of a pro player" for game League of Legends.
Top 10 Most Anticipated Video Games of 2020
2020 will have something to satisfy classic and modern gamers alike. To be eligible for the list, the game must be confirmed for 2020, or there should be good reason to expect its release in that year. Therefore, upcoming games with a mere announcement and no discernible release date will not be included.
Top 15 NEW Games of 2020 [FIRST HALF]
2020 has a ton to look forward to...in the video gaming world. Here are fifteen games we're looking forward to in the first half of 2020.