Decided to start a new thread since the thread http://pinside.com/pinball/forum/topic/dont-look-now-but started becoming more of a which game is best.
I love pinside but I also interpret the ranking with a grain of salt.
Tron #1 over MM.
I have done a ton of analytical work including algorithms for rating systems. It is a very slippery and complex slope. I do think the Pinside ratings could be improved but I am thankful we have a site like this.
When dealing with number ratings there are a ton of variables to account for. The first and most simplest is the “Trimmed Mean” which basically throws out high and low that can skew. But this does not account for times ranked which is dependent on who created the ranking system and how they want it to be used.
Some sites do not care to count the times ranked and when they do they limit it to a mininum times ranked. But this does not yield 100% accurate results.
For this site
27 votes for Tron and 695 for MM.
Just hypothetical Tron could have gotten 27 votes above 9.00.
MM could have gotten 300 votes above 9.00 and 100 votes at 10.00 and then maybe 200 below 8.00, etc, etc.
So even though MM has more 9.00 and 10. Votes than Tron it will still rank lower due to the outliers, etc.
To me a ranking means nothing without its TOTAL Rankings applied to outcome. So to me Tron at #1 with 27 votes means nothing to me.
But this is just me... some sites like Amazon apply the ranking regardless of the times reviewed.
Really boils down to with 27 votes and 695 for MM you are not comparing Apples to Apples.