[5 maj 13]

This is a zestfully amazing thread, and I apologize for asking an unrelated question...but, have any of you found it difficult if not impossible to render an opinion contrary to management anywhere on the site?

For example, I've never had a TCR below 90 until recently I upset one of the goulash hosts, at which point my TCR rating dropped from 90% to 50% overnight. At first I thought it was some sort of error. At last, it's an obvious sandboxing set by hand by some admin...completely sinister among those of us who play a game of honor...not a percentage rating at all, but a number assigned by some toadie, some friend of the goulash cheerleaders.

BBO, PLEASE DON'T CONTINUE PRETENDING THE TCR IS A PERFORMANCE METRIC. IT'S A TOOL YOU USE TO REWARD AND PUNISH THOSE WHO QUESTION YOUR AUTHORITY.

Easy enough, it's true.

I deeply regret the 5000+ logins I've logged here, and especially the money invested in (what I now realize to be bogus) BBO tournaments to earn a GOLD 13 by my name.

I'm utterly disgusted. I'll never contribute another penny to this lawless, bogus website. I'll rejoin those who play for free and suspect the worst of the professional card cheats who seek our nickels and dimes.

Best regards,
5 maj 13

[barmar]

TCR cannot be set manually, it's calculated automatically every day, and there's no way to override it. Upsetting a tourney host will not affect it.

A reason why it can change suddenly is because we only recalculate TCR when you have at least 10 tourneys in the preceding 60 days. If you don't play in tourneys for several months, your TCR will stay the same. If you then start playing tourneys again, it will not change until you have 10 recent tourneys. After the tenth tourney, it will be recalculated based on your percentage of those tourneys that you completed.

It can also change significantly because we only consider the tourneys in the previous 60 days. So consider something like this:

Day 1: Play and complete 20 tourneys
Days 2-59: Play 10 tourneys and complete 8 of them.

At this point you have a 93% TCR.

When we recalculate the next day, the 20 tourneys on day 1 are no longer included, and your TCR drops to 80%.

I can't easily tell precisely what happened in your case. That goulash tourney was over a month ago, and we only keep 90 days of tourney completion history. So I can't tell what happened in the full 60 days leading up to it.

You've only played 6 tourneys in the past 60 days, so your TCR is stuck there. 3 of them were played almost 60 days ago, so they'll age out of the calculation in the next day or so. Of the 3 more recent tourneys, you only finished 1. So you'll need to play and complete a bunch of tourneys in the next 2 weeks to get your TCR recalculated and improve your percentage significantly.

You could argue that our method is too simple, but there's no malice or prejudice involved. As I said above, there's no way for a tourney host to change your TCR.

[pescetom]

barmar, on 2019-December-27, 09:38, said:

You could argue that our method is too simple, but there's no malice or prejudice involved. As I said above, there's no way for a tourney host to change your TCR.

The ideal would be to age each data point with declining weight over time. But a simple modification sufficient to make it behave a bit more intuitively would be to recalculate only a certain part (say 50%) based on the last 60 days.

[barmar]

There's value to simplicity -- if someone asks why they have a particular TCR, I can show them their history and the calculation can be done easily in your head.

Remember what the purpose of TCR is -- it's to estimate how reliable you and your Internet connection are. This isn't something that tends to degrade or improve gradually, so that an exponential decay makes sense. We just assume that your recent history is representative, and we picked 2 months and 10 tournaments as reasonable parameters to get a useful approximation.

[pescetom]

barmar, on 2019-December-29, 14:46, said:

There's value to simplicity -- if someone asks why they have a particular TCR, I can show them their history and the calculation can be done easily in your head.

Remember what the purpose of TCR is -- it's to estimate how reliable you and your Internet connection are. This isn't something that tends to degrade or improve gradually, so that an exponential decay makes sense. We just assume that your recent history is representative, and we picked 2 months and 10 tournaments as reasonable parameters to get a useful approximation.

I'm with you, but that's why I suggested 50% - half me, half my recent internet connection.
Some incompletions are internet related, some are player choices - it's the latter you are trying to discourage, so I think you should cut some rope for those who have always finished but are temporarily on a bad connection.

[barmar]

pescetom, on 2019-December-29, 15:56, said:

I'm with you, but that's why I suggested 50% - half me, half my recent internet connection.
Some incompletions are internet related, some are player choices - it's the latter you are trying to discourage, so I think you should cut some rope for those who have always finished but are temporarily on a bad connection.

We don't really care why you drop out. It disrupts the tournament either way, so we're not trying to "discourage" anything. But we generously assume that it's mostly due to bad connections, which is why we think that recent history is predictive.