Conversion Score

[JasonQiu21]

Is your feature request related to a problem? Please describe.
Conversion rate is a rather flawed statistic in that it values tops of all tournament sizes equally; conversion score is an attempt to fix this; credit to datatog for finding this

Conversion Rate is topCuts / tournaments
Conversion Score is calculated as follows:

• Calculate the 'weight' of each tournament on the assumption that your chances of making top cut are equal to all competitors - this is sum(topCut/tournamentSize); call this expectedTops
• Your conversion score is then topCuts/expectedTops - anything above 1 is better than expected; anything below is worse than expected
• I like it multiplied by 100 so it resembles plus-statistics in baseball like wRC+, OPS+ (even though 100 is not average, rather it's expected)

Describe the solution you'd like

Instead of conversion rate, display conversion score by default and include a toggle to old CR (or not - if it's flawed, do we even want to display it?)

Additional context
#28 (comment)
https://www.reddit.com/r/CompetitiveEDH/comments/17w5exx/edhtop16_conversion_rates_have_a_mathematical/

[rrdelaney]

Helpful email I received this morning:

Hi,
I am reaching out to you to ask about your plans of correct the issue with Conversion Rate giving out misleading information.
Conversion Rate is a tournament specific metric, and it can’t be used in aggregate with other tournaments.

It’s calculated on the idea of what is the probability of a specific deck making it into the top cut, and since the tournament sizes vary, they can’t be combined for a “General Conversion Rate” like you have done.

For example, if a tournament has 500 participants and 160 of them are on for example Blue Farm, even with everyone in the top 16 being Blue Farm, the Conversion Rate would still only be 10%.
That 10% is the best the deck could do, since there are only 16 spots I the top 16, naturally.

If a tournament on the other hand had 100 players, and 32 of them were on Blue Farm, a top 16 of only Blue Farm decks would mean a Conversion Rate of 50%.
Nothing about the performance of the deck changed, but due to the flaws of the Conversion Rate, they give vastly different results.

By mashing together these two conversion rates, like you do on your site, you get a “non-value” of 30% that does not tell anything about the deck’s performance.

I suggest you redo the way you measure performance. A much better way would be to calculate
Population in top 16 / Population in the tournament.

That way the first tournament would give a result of 16 / 16 : 160 /500 = A performance of 3.125
And the second would give a performance of 16 / 16 : 32 / 100 = 3.125

Those “performance values” can be averaged out and the average value is more useful for the purpose of gauging power. I hope you will fix this I the future, since many of the people playing the format don’t understand the problems with using Conversion Rate as a metric, and I feel like you are misleading them by stating that information.

Br.
Jonne Välimäki, from Finland