Competition Analytics - Atlantic City

[BMinahan]

As some of you know, I run analytics on our competitions to see what went right and most importantly, what went wrong. Based on this discussion, I decided to share my analysis of this past competition in AC. The basis of the analytics is to determine what tables were good tables and which were bad. Based on that data, I can analyze our cook compared to the table rankings. I am not going to go into detail on how the software ranks each table in this forum but if you are interested, please see me at a competition and I would happily explain. In addition, the analytics are built for teams that consistently finish in the top 20% of the field. If your team does not typically do that, then this analysis will not help you. This is designed to help dial in your recipes once you hit the top 20%. So here’s the rub.

#1. My analysis of the cook. I thought all of our categories were on point. Especially the chicken, ribs, and pork. The brisket was slightly over however I felt the slices we turned in were good enough for a top 10 call. The burnt ends rocked! In comparison to other competitions we have won, this was a really good cook. Our expectation was a top 5 finish with a shot at the GC.

#2. Our scores: 8th Chicken (TBL#327), 34th Ribs (TBL#646), 5th Pork (TBL#981), 35th Brisket (TBL#642), Overall 22nd.

#3. Ranking the tables. After running the data through my software here is what I found.
Best tables to hit (in order): 327, 724, 315, 311
Good tables: 981, 661, 650
Tables that won’t kill you as long as you hit one of the best tables: 449, 526
Bad Tables (really hard to win if you hit these. Need to hit 327 & 724 to make of for this deficit): 977, 124
Tables of Death (no chance of winning): 642, 646

Percentiles: The top possible percentile was 79% which is made up of the top 4 tables. The bottom percentile was 8% if you hit the bottom 4 tables. The top 20 percentile was 57%. This means that in order to have any shot at winning, you needed a score higher than this. Based on the 15+ contests I ran through my system, this has always been the case.

Que and a Half Men (my team): 35%. We hit the best table for Chicken, a middle of the road table for Pork, and both tables of death… Put a fork in us, we had no luck in AC.

ZBQ: 66%. They hit the best table for Brisket, and good tables for Pork and Chicken. They also hit 526 for their Ribs but we able to offset that with the Brisket table.

Dueling Jambos: 71% Hit two top tables for Chicken and Brisket and good tables for Pork and Ribs. They had a good cook with some great luck in AC!

Insane Swine: 61%. Had to run Jared’s numbers as he scored a 700+ score in Bentonville. It is really hard to comprehend how his score dropped to 648 is just two weeks. I am sure his cook did not change much. Jared hit the top 2 tables for Chicken and Brisket and good tables for Pork and Ribs. Not too sure what happened there.

Dr. Pearls: 36% Hit a top table for Brisket, a good table for Ribs, hit a bad table for Chicken, and the death nail was Pork on table 642.

I hope this helps make some sense of the luck factor in competition BBQ!

 

[INmitch]

Interesting. Give me 10-15 minutes looking over my score sheets and I pretty much get a general picture of what went on. But that nails it down. I like it.

 

[Pappy Q]

I'd buy that in an app

 

[midwest_kc]

I'm curious as to why it won't help you if you're not consistently in the top 20%? I have my ideas (you still aren't cooking it properly, and have issues with presentation), but I'm curious your thoughts.

 

[TheJackal]

Bill- Do ours. I'm a math major and work for a software company that specializes in Data Integration and Analytics (including Data Quality, Master Data Management, Data Governance, etc). I always analyze scores. I place them in a spreadsheet and look at good and bad tables as well. Mark took the sheets and hasn't faxed them to me yet so I don't recall the table numbers we hit (or who else was on them). I'd be interested in seeing your analysis based on what I think (emotionally) happened vs what the numbers tell.

Chicken: It was great chicken. I expected a top 5. I assume it hit a table other than the TODs

Ribs: 4 calls in the past 5 contest. They were really good but not our best. Assume it hit a table other than the TODs.

Pork: Very erratic category for us. Good box and good tenderness. We are finding our way on taste. I assume it hit one of the better tables.

Brisket: It was not great and would need help from a TOA. I assume it did not hit one of the better tables.

Whatcha got for me?

 

[CBQ]

I thought there were some interesting insights here. It's sort of what we do manually by looking at KCBS score, but by applying analytics to it you can see more. I like the handicapping of teams based on how high or low scoring the tables were - did you have good luck, bad luck, or judging that truly reflected what you cooked?

What we can't see is what really hit the table. Was a table a high scoring table because the judges were generous, or did a bunch of top teams hit the same table? (In other words, judges can also be lucky or unlucky, and the table score will reflect that.)

If we could take judges scoring across a larger sample (multiple contests) you could do a more accurate assessment of the "luck" factor. (As I have said before, I would like to see KCBS use data like this to determine seating of judges at a contest - blending judges to reduce the "table of death" effect.

 

[joemat]

So AC was only my third competition cook (previous cooks were Q for the Troops and Staten Island this year). My best result this weekend was 13th in Pork from table 124 (bad table). So I was happy to walk on that. Unfortunately it went downhill from there. Ribs were at the table of death (642):sad: ...I get the drift.

Still can't get the hang of Chicken for some reason but I guess I have the winter to figure that out.

Congrats to ZBQ and Dueling Jambos!!

 

[TheJackal]

What we can't see is what really hit the table. Was a table a high scoring table because the judges were generous, or did a bunch of top teams hit the same table?

We can't see that one entry in that specific contest but what I gathered from Bill's '20%' comment is he takes this into account over time. So if you are a 20% team and scored 1st on your table, but wound up on a table with 5 other non-20%-ers. Guess what? Probably wasn't a TOD and you just had a bad day. It happens.

Same goes for the reverse. If as the only 20%-er, you were 6th in a category but 3rd on your table, you probably hit a TOA.

Now if I was 6th in a category and 3rd on the table, and 1 and 2 were Que and a Half Men and ZBQ, then I might call it a 'bad' table to land on :biggrin1: but it wasn't a TOD and may not have been a TOA either.

:decision::decision::decision::decision::decision::decision::decision:

 

[CBQ]

Now if I was 6th in a category and 3rd on the table, and 1 and 2 were Que and a Half Men and ZBQ, then I might call it a 'bad' table to land on :biggrin1: but it wasn't a TOD and may not have been a TOA either.

Ya, that's the thing you can't infer from this...if you happen to hit spot 6 on a table and have a bunch of heavy hitters in front of you that might not be good.

 

[TheJackal]

Can sites like BBQscores.com get table numbers with their data extracts? That would open up a whole new avenue of reporting and analysis. I wish KCBS would start listing all score sheet info in their online results. They have the data. Why not post it?

 

[71-South]

Interesting stuff. Right now, my app just takes as input the individual judges' scores for the 4 meats and the rank in each category (nth out of N). I'm trying to keep the data entry at a minimum while getting as much analysis as possible. I think I'm going to add a place to put the rank at each table to come up w/ a poor man's look at TOAs vs. TODs. Here's my thought. Basically, if you finish 1st at your table, you could reasonably assume you'd come in somewhere in the top 6th in that category. If you finished 2nd at your table, you could assume you'd come in somewhere in the 2nd 6th in that category. Etc. If your ranking on the table differs drastically from your overall ranking in the category, that would be a red flag (or a green flag?)

I'd love to hear thoughts on that from people much mathier than me.

I may add in places to put much, much more data, but just not make it mandatory. We'll see how it goes.

 

[markG]

I think that the top 20% deal may be that if you team is in the Top 20% of a category or overall in most competitions that you compete in, you probably have your cooking technique and recipes dialed in, so you can judge if you hit your marks. Someone that cooks 2 or 3 times a year (with the exception of Tuffy Stone) may not have a finger on the pulse of what they are making, so it is hard to judge what you see as good vs. what the judges see.

 

[markG]

a downloadable xls or cdf from KCBS would be nice to have. I bet it took quite a bit of time for Bill to get the numbers into his database. Nice work Bill!

 

[BMinahan]

Smokopolis Analysis

Percentage: 55%
Really close to being in the window for a GC.
Chicken - Hit the 2nd best table and you won the table. Cant do much better there.

Ribs - Hit a good table. You won the table and they rewarded you with one of their three (all categories) top 10 calls.

Pork - Hit the 4th best table against Chris Hart. We know that boy can cook! You earned 10th place so that is good... You and Chris tied on judge 1 but you lost on 2,4, & 6. 3 & 5 liked yours better. Check the appearance, taste, tenderness scores to see where there could possibly be a minor tweek for improvement.

Brisket - a bad table. This one has me perplexed as you got 3rd on the table and they crushed you on the scoring.

 

[BMinahan]

Mark G - Right on.

 

[BMinahan]

Mark G - Yes. It is quite a bit of data entry and I always worry about keying errors.

 

[TheJackal]

Thanks, Bill! Brisket has me perplexed too. lol.

 

[BBQchef33]

As some of you know, I run analytics on our competitions to see what went right and most importantly, what went wrong. Based on this discussion, I decided to share my analysis of this past competition in AC. The basis of the analytics is to determine what tables were good tables and which were bad. Based on that data, I can analyze our cook compared to the table rankings. I am not going to go into detail on how the software ranks each table in this forum but if you are interested, please see me at a competition and I would happily explain. In addition, the analytics are built for teams that consistently finish in the top 20% of the field. If your team does not typically do that, then this analysis will not help you. This is designed to help dial in your recipes once you hit the top 20%. So here’s the rub.

#1. My analysis of the cook. I thought all of our categories were on point. Especially the chicken, ribs, and pork. The brisket was slightly over however I felt the slices we turned in were good enough for a top 10 call. The burnt ends rocked! In comparison to other competitions we have won, this was a really good cook. Our expectation was a top 5 finish with a shot at the GC.

#2. Our scores: 8th Chicken (TBL#327), 34th Ribs (TBL#646), 5th Pork (TBL#981), 35th Brisket (TBL#642), Overall 22nd.

#3. Ranking the tables. After running the data through my software here is what I found.
Best tables to hit (in order): 327, 724, 315, 311
Good tables: 981, 661, 650
Tables that won’t kill you as long as you hit one of the best tables: 449, 526
Bad Tables (really hard to win if you hit these. Need to hit 327 & 724 to make of for this deficit): 977, 124
Tables of Death (no chance of winning): 642, 646

Percentiles: The top possible percentile was 79% which is made up of the top 4 tables. The bottom percentile was 8% if you hit the bottom 4 tables. The top 20 percentile was 57%. This means that in order to have any shot at winning, you needed a score higher than this. Based on the 15+ contests I ran through my system, this has always been the case.

Que and a Half Men (my team): 35%. We hit the best table for Chicken, a middle of the road table for Pork, and both tables of death… Put a fork in us, we had no luck in AC.

ZBQ: 66%. They hit the best table for Brisket, and good tables for Pork and Chicken. They also hit 526 for their Ribs but we able to offset that with the Brisket table.

Dueling Jambos: 71% Hit two top tables for Chicken and Brisket and good tables for Pork and Ribs. They had a good cook with some great luck in AC!

Insane Swine: 61%. Had to run Jared’s numbers as he scored a 700+ score in Bentonville. It is really hard to comprehend how his score dropped to 648 is just two weeks. I am sure his cook did not change much. Jared hit the top 2 tables for Chicken and Brisket and good tables for Pork and Ribs. Not too sure what happened there.

Dr. Pearls: 36% Hit a top table for Brisket, a good table for Ribs, hit a bad table for Chicken, and the death nail was Pork on table 642.

I hope this helps make some sense of the luck factor in competition BBQ!

Welp, this eases the sting a little bit.. I finally found my score sheet and checked.. We came in 42/65 overall.. probably the worst finish ive had in years. i had expectations for chicken and brisket, and maybe a shot at pork.. but....

Chicken hit a good table 327 and came in 26th.. Thought it was better than that..

We thought our brisket was dead on and going to hit at least top 10, as did several judges that tasted it after the contest.

BUT... our brisket hit the table of death on 642 earning 48th with all 7'and 1 8. I was pretty shocked.

pork ribs hit the bad table on 124 for a 39th. yeah, they sucked.. i didnt have any expectations for them.

pork hit meh table at 526 for 31st..

goes to show theres still the luck of the draw in hitting the good tables in our equation. but.... no big deal.. we had a great time.. we were new team that never cooked together and it all ran smooth.. used some new rubs, new techniques.. no complaints, no regrets.. we'll be back again. .

 

[goodsmokebbq]

You just got to get good enough to get lucky

 

[BBQchef33]

You just got to get good enough to get lucky

nope.. not me.. no matter how good I get... lady luck avoids me even worse than fish... which btw... I havent caught any this year and its not for lack of trying. :mmph:

Table of Death.. here we come!! :laugh::laugh::laugh: