Hey Jace,

Does the expression "TGSAP error rate" refer to voter accuracy, or something else?

I posted a simple adjudication summary 2 years ago. It probably is NOT what you're looking for, given that it only covers a 10-day period in November 2017 as opposed to cumulative results (link below):

November 2017 Adjudication Summary - 10 Days

Jace Hall thanked this post

datagod liked this post

The post he is referring to is a post that he believes was made within the last 3 weeks.

I can't seem to find it either.

I am referring to a recent comment that was in a wall post within the last 2-3 weeks. Someone looked at the total number of adjudications, and disputes and etc. and calculated an effective error rate (something like .07%) or something... @stella_blue do you remember seeing a comment like that? Perhaps @The Evener saw it?

Originally Posted by Jace Hall

I am referring to a recent comment that was in a wall post within the last 2-3 weeks. Someone looked at the total number of adjudications, and disputes and etc. and calculated an effective error rate (something like .07%) or something... @stella_blue do you remember seeing a comment like that? Perhaps @The Evener saw it?

I scanned every wall post over the past 2 months (11-28-2019 to 01-28-2020), drilling down on any discussion that seemed promising.

Nothing so far.

My guess is that the comment has since been deleted, or is buried among other comments under a misleading thread title.

The closest thing I've found is this entry by Andrew (Barthax):

https://www.twingalaxies.com/barthax/wall/7706/useless-stats-from-the-pulse-of-the-first-4-years-of-tgsap

Some interesting stats, but no mention of a TGSAP error calculation.

Hmm...

I remember the comment being on a post that was not specifically on the topic of stats.

I can’t fully remember but I have a feeling the subject had something to do with somebody’s CR being impacted or how blind voting was a problem and that the comment said that it wasn’t. And I just can’t remember but it was within the last three weeks or so...

Ughghh

Thanks for trying to help everyone. It may just be a lost comment.

Originally Posted by Jace Hall

Hmm...

I remember the comment being on a post that was not specifically on the topic of stats.

I can’t fully remember but I have a feeling the subject had something to do with somebody’s CR being impacted or how blind voting was a problem and that the comment said that it wasn’t. And I just can’t remember but it was within the last three weeks or so...

No, you're right, I'm pretty certain it had to do with blind voting. I feel like I remember this too.

Jace Hall liked this post

Eureka!

https://www.twingalaxies.com/entry.php/7693/7693-Pausing-Is-it-allowed

stella_blue thanked this post

@stella_blue calculated individual success rates using TGSAP which can be found here

https://www.twingalaxies.com/ninglendo/wall/5774/weeding-out-blind-voters/page/2#comment33529

SincerelyFranny, GibGirl thanked this post

Originally Posted by Jace Hall

Hmm...

I remember the comment being on a post that was not specifically on the topic of stats.

I can’t fully remember but I have a feeling the subject had something to do with somebody’s CR being impacted or how blind voting was a problem and that the comment said that it wasn’t. And I just can’t remember but it was within the last three weeks or so...

Ughghh

Thanks for trying to help everyone. It may just be a lost comment.

Hey Jace,

The comment you're looking for was posted on 01-14-2020 by Kyle Nelson (thegamer 1185) :

https://www.twingalaxies.com/thegamer1185/wall/7693/pausing-is-it-allowed/page/6#comment52955

Originally Posted by thegamer1185

There have been 59,645 TGSAP accepted scores and 314 disputed scores accepted with another 130 being disputed. Many of those disputed scores are not TGSAP related either. That equals .007% of ALL scores in the database. TGSAP is apparently working quite well. I suppose some of those 59K TGSAP scores could be incorrect, but I think there are more voting properly than there are blind voters. I'd rather see the rules updated than worry about blind voters affecting anything.

A special thank you to Franny for directing me to the relevant thread.

SincerelyFranny, GibGirl, Jace Hall thanked this post

Originally Posted by Ninglendo

@stella_blue calculated individual success rates using TGSAP which can be found here

https://www.twingalaxies.com/ninglendo/wall/5774/weeding-out-blind-voters/page/2#comment33529

I remember performing that analysis, but forgot that I actually posted it somewhere.

Whoops!

SincerelyFranny liked this post

@thegamer1185 said:

There have been 59,645 TGSAP accepted scores and 314 disputed scores accepted with another 130 being disputed. Many of those disputed scores are not TGSAP related either. That equals .007% of ALL scores in the database. TGSAP is apparently working quite well. I suppose some of those 59K TGSAP scores could be incorrect, but I think there are more voting properly than there are blind voters. I'd rather see the rules updated than worry about blind voters affecting anything.

While I am sure that a deeper dive into this would yield greater detail and accuracy, I just want to say that until I saw this statement I truly did no realize how well TGSAP has been working.

Most of the time I am focused on dealing with the complaints and etc. as it relates to the system and so it can begin to feel like the issues are overtaking things, but it was refreshing to be able to take a step back and catch a glimpse of the bigger picture.

This is truly an astonishing success rate and it would be somewhat surprising to me to find that any other video game performance tracking entity matches this kind of adjudicative accuracy consistently.

I think this is something for the TG community to be proud of and to help others understand what it means to participate at TG and why it takes the approach that it does.

Everyone here really makes an important contribution. Not just to TG, but to the video game community at large. In my opinion, over time, the impact of what is happening here will be felt more and more.

Thank you to everyone for helping to find this comment, and huge thanks to @thegamer1185 for taking the time to point it out in the first place!

Garrett Holland, thegamer1185, SincerelyFranny, Barthax liked this post

Originally Posted by Jace Hall

@thegamer1185 said:

While I am sure that a deeper dive into this would yield greater detail and accuracy, I just want to say that until I saw this statement I truly did no realize how well TGSAP has been working.

Most of the time I am focused on dealing with the complaints and etc. as it relates to the system and so it can begin to feel like the issues are overtaking things, but it was refreshing to be able to take a step back and catch a glimpse of the bigger picture.

This is truly an astonishing success rate and it would be somewhat surprising to me to find that any other video game performance tracking entity matches this kind of adjudicative accuracy consistently.

I think this is something for the TG community to be proud of and to help others understand what it means to participate at TG and why it takes the approach that it does.

Everyone here really makes an important contribution. Not just to TG, but to the video game community at large. In my opinion, over time, the impact of what is happening here will be felt more and more.

Thank you to everyone for helping to find this comment, and huge thanks to @thegamer1185 for taking the time to point it out in the first place!

Just to clarify, the error rate was misprinted in Kyle's assessment. The true error rate is 0.7%, not 0.007%, (444÷59,645=0.007444... or 0.7%,) which is still exceptionally successful.

Jace Hall thanked this post

Originally Posted by Garrett Holland

Just to clarify, the error rate was misprinted in Kyle's assessment. The true error rate is 0.7%, not 0.007%, (444÷59,645=0.007444... or 0.7%,) which is still exceptionally successful.

A 1% or less error rate is pretty incredible from my perspective.

Garrett Holland liked this post

59,977 submissions accepted

3,570 rejected

94.38% of all non canceled submissions are accepted. This percentage is likely light, as rejected subs can also be rejected rather than canceled for a number of reasons (cancellations are no longer an option comes to mind).

While this stat doesn't directly relate to voting accuracy, it does reflect that most subs are accurate at the time of submission.

Jace Hall liked this post

Originally Posted by Jace Hall

A 1% or less error rate is pretty incredible from my perspective.

... I'll go further and say that even "incredible" is an understatement. If my company had an error rate even remotely close to that, I'd be very wealthy.

Jace Hall liked this post

Originally Posted by Garrett Holland

Just to clarify, the error rate was misprinted in Kyle's assessment. The true error rate is 0.7%, not 0.007%, (444÷59,645=0.007444... or 0.7%,) which is still exceptionally successful.

Ha, you are right. I have been using an Excel sheet that has been multiplying by percentages which appear as .01 for 1%, 1.00 would be 100%. So in Excel format I was correct :P