PDA

View Full Version : Alternative Data to TOY? Average Score per Comp?


Teamfour
09-26-2013, 08:30 AM
Is there anyone tabulating the average score for each team for overall and category points? Talking about KCBS national comps.

I think the TOY rankings are great and kudos to those teams at or near the top. But we all know that a big part of the TOY rankings are the number of comps a team enters.

As a new team, I would be more interested knowing where I stand related to the average score for each team - total points divided by number of comps. I think it would be interesting to measure consistency as compared to the top dogs.

Comments?

Slamdunkpro
09-26-2013, 08:33 AM
With regional variance in scoring it wouldn't be a level playing field unless you factor it.

Teamfour
09-26-2013, 08:35 AM
With regional variance in scoring it wouldn't be a level playing field unless you factor it.

Thanks Michael. I edited my post to specifically target KCBS national comp data. I think the scoring is standard across all KCBS comps.

Nordy
09-26-2013, 08:42 AM
Took me a minute, but I think I gotcha...

You want to use the TOY points each team earns per contest and average them Overall, and Categories...

Interesting... That takes regional score bias out of the picture....

Unless you use ALL competitions (not just top 10 comps like TOY does) I think you will find the results are similar, especially for teams with 10 or more comps. However it gets a little interesting if you use all 25-35 comps that some teams do and not just 10...

N

Slamdunkpro
09-26-2013, 08:43 AM
Thanks Michael. I edited my post to specifically target KCBS national comp data. I think the scoring is standard across all KCBS comps.

But it isn't. For example the average winning scores in the north east are considerably lower than those in the mid west. A 660 might win an event in the northeast and be 50th in the mid west. Look at Sams in Medford for an example.

Teamfour
09-26-2013, 08:51 AM
Unless you use ALL competitions (not just top 10 comps like TOY does) I think you will find the results are similar, especially for teams with 10 or more comps. However it gets a little interesting if you use all 25-35 comps that some teams do and not just 10...



Correct. I want to see the average of ALL comps entered.

Teamfour
09-26-2013, 08:56 AM
But it isn't. For example the average winning scores in the north east are considerably lower than those in the mid west. A 660 might win an event in the northeast and be 50th in the mid west. Look at Sams in Medford for an example.

Well, I'm already hijacking my own thread, but...

I see what you are saying. However that is an indication of the regional difference in judging, not necessarily the quality of the team's cooking.

The KCBS TOY system assigns points for where you place plus a factor for the number of teams in the comp. So the actual judge scores are substituted with a score for where you place regardless of the raw score. I think this takes the regional judging trends out of the picture.

Slamdunkpro
09-26-2013, 09:07 AM
Slow pony here - I get it.

cpw
09-26-2013, 09:19 AM
bbqscores.com will show you your own average in relation to a regional average for each category, per team. Looks like you'd have to compare your own scores with other teams to see where you stand. For example, our average Brisket score is 167.9923 with 3 contests cooked, in comparison to Chatham Artillery's average of 171.4943 with 8 contests (I chose Chatham because he competes in the same region).

Kit R
09-27-2013, 08:29 AM
Correct. I want to see the average of ALL comps entered.

Interesting idea as it factors out the regional scoring variance. The next, and unfortunately necessary, question is how many comps is enough to get a representative sample of scores to insure the calculated average is meaningful? I hate to say it, every time I see a great idea like Lee's proposed I immediately think of the way(s) people will try to game the system. Like find three "bottom feeder" events, beat the competition like a drum and pump up your team average? And once your average was high enough, you could just stand pat and not risk lowering it at another coimp with a more, er, robust field of competitors.

Like I said, I think Lee's idea has some merit. But to the extent possible, you'd have to think through all the scanarios for manipulation and gaming and address them. Until the next loophole was discovered and exploited, that is...........

Dan - 3eyzbbq
09-27-2013, 10:59 AM
A quick sniff test and it doesn't work. Our TOY point total is 2720/22 comps = 123(ish)

TrueBud is 2938/29 comps is 101

A quick glance at the TOY standing and you see Truebud has the lead.

Am I missing something here?

Kit R
09-27-2013, 11:12 AM
A quick sniff test and it doesn't work. Our TOY point total is 2720/22 comps = 123(ish)

TrueBud is 2938/29 comps is 101

A quick glance at the TOY standing and you see Truebud has the lead.

Am I missing something here?

If you calculate the averages on your respective "best 10" (current KCBS TOY criteria), how do they come out?

Dan - 3eyzbbq
09-27-2013, 11:29 AM
Rhetorical question Kit?

He was talking about averaging ALL comps, was just pointing out that doesn't work.

Jorge
09-27-2013, 11:38 AM
A quick sniff test and it doesn't work. Our TOY point total is 2720/22 comps = 123(ish)

TrueBud is 2938/29 comps is 101

A quick glance at the TOY standing and you see Truebud has the lead.

Am I missing something here?

I'd guess that something is off there. I'd guess that 2938 is a 10 contest point total. 250 pts. per win, and additional points (maxed at 50 per) for each contest. Even then I'm not sure the #s are right. Taking their top 10 wins I'm surprised that they lost 62 pts. from the team count.

I'd expect their top 10 to be closer to 3000 than the number quoted.

Kit R
09-27-2013, 11:42 AM
Rhetorical question Kit?

He was talking about averaging ALL comps, was just pointing out that doesn't work.

No not at all. I was trying to say that the two different criteria yield different results. If you average them all in , it's going to be two different outcomes for sure.

TheJackal
09-27-2013, 01:11 PM
A quick sniff test and it doesn't work. Our TOY point total is 2720/22 comps = 123(ish)

TrueBud is 2938/29 comps is 101

A quick glance at the TOY standing and you see Truebud has the lead.

Am I missing something here?

I don't think that's what the OP is intending. 2720 is a 10 comp point total. He wants to take the total of all 22 comps and then average it. So yours would be 4914/22=223.36

TrueBud would be 7027/29=242.31

Ours would be 1294/7=184.86

Essentially I think he's looking to comparatively rank teams that do 10+ comps against those that do less than 10.

Dan - 3eyzbbq
09-27-2013, 01:26 PM
Thanks Greg - thats what I was missing :)