Case Study: A Trend Analysis of the "Trap Game" Scenario

[JROCK1966]

Introduction

First, I'll state several quick background facts about me. One, I'm a newbie when it comes to SDQL. Two, I think I am a quick learner though. Three, this write-up concerns the Old Dominion / FIU game from Nov. 11th, 2017. Four, I had a gut feeling this line was off when I first saw it. Please refer to post #3 of the Week 11 ML Dog thread where I stated as much. Five, only after the result did I have the think to investigate this game further. Six, while hindsight thinking is 20/20, I now know how to identify these types of games in the future.

 

[JROCK1966]

Secondly, it was not my use of SDQL that was so much a problem when I applied it last week. It was my train of thought on how to approach the research. In general, I was trying to find dogs last week that had stats in their favor thinking I can identify games that had a lot going for the team getting points. This line of thinking / thought process / whatever you want to call it I now know is wrong. It's a "public" way to approach capping. Good luck finding those games. For the most part, my results came back 50% which is not very helpful. What I need to think about in the future is the exact opposite. Find either 1) dogs that have almost nothing going for them and the line looks off and/or 2) favorites that are doing great, have a lot going for them, and the line seems very reasonable. This line of thinking lends itself to the ages old "Trap Game" scenario we have all heard about, experienced, but that I have never been able to recognize and prove it exists. Until now that is. With this in mind I played around with some scenarios this am and here is what I found:

 

[JROCK1966]

Case Study: ODU @ FIU, November 11, 2017:

A few facts heading into this game....FIU had just beaten Marshall on the road and UTSA at home and was dogged in both games. This led me to believe they should handle the "reasonable" -7 line against a poor ODU squad. I did research this game but concluded there were no stats to support such a play on ODU. In short, I was applying the "public" line of thinking towards capping this game. I mean, after all, ODU was more turnover prone, total defense was worse, and their QB efficiency was worse. Nothing from those 3 stat categories suggested an ATS cover and certainly not an upset win on the road....or did they??

 

[JROCK1966]

Let's first look at the historical ATS and SU record of dogs that had a line of >10pts. The historical record is BAD with a win % of 8% and an ATS cover rate of 44.8%. Please see the green boxes in the pic.

 

[JROCK1966]

Next let's isolate the historical ATS & SU records for just the years 2016 & 2017. Again, the historical record is EVEN WORSE with a win % of 6.3% and an ATS cover rate of 42.6%. Please see the green boxes in the pic.

 

[JROCK1966]

Next, let's further isolate the historical ATS & SU records for just the year 2017. Again, the historical record is EVEN WORSE with a SU win % of 4.3% and an ATS cover rate of 39.1%. Please see the green boxes in the pic.

 

[JROCK1966]

With all those terrible results listed in the above posts, one might think that this scenario would be consistent across the board, right? Well, no. The answer is not only no, but hell no! Listed below are results of my SDQL queries for lines less than or equal to 10pts. This scenario perpetuates the thought process I had, when I first looked at the line and said, "WTF? Why is FIU only laying 1 score against ODU?" Hence, the ages-old "trap-game" scenario or "sucker-bet" as otherwise known as.

You'll see the historical results for dogs, where the line is less than 10pts, are better.....not great but better with a 41% SU win % and 55% ATS win%....but wait, there's more...

 

[JROCK1966]

For some reason I do not know why yet, these dogs the past two seasons are hitting ATS and SU at a remarkable rate. This year alone, this scenario is hitting SU at 61% and ATS at 77%.

 

[JROCK1966]

The records for the 2016 & 2017 seasons combined are 66.7% SU and 83% ATS...Wow!

 

[JROCK1966]

Conclusion:

Yes Charlie Brown! There is such a thing as a "trap game" or "sucker-bet" scenario. Be very leery of backing a favorite who seemingly can cover a perceived low, reasonable line. I'm hoping to find a few of these as the 2017 season winds down and also will try to apply it to other sports as I build my SDQL knowledge level. Peace out my friends!

:shake:

 

[booksbestfriend]

Can you put in conference and away team being favorite in your query?

 

[JROCK1966]

Can you put in conference and away team being favorite in your query?

The exact opposite you mean? Sure, give me a few minutes.

 

[JROCK1966]

never mind I re-read your post...stand by...

 

[JROCK1966]

Can you put in conference and away team being favorite in your query?

Per your request....historical records for road conference favorites, line > 10pts, better QB efficiency, both teams positive turnover margins.

Numbers appear to be what you would expect.

 

[JROCK1966]

Historical records for road conference favorites, line between 0 and -10, better QB efficiency, both teams positive turnover margins...not quite as good as historical records for lines larger than -10.

 

[JROCK1966]

2017 records with line between 0 and -10:

 

[JROCK1966]

2016 & 2017 records with line between 0 and -10:

 

[booksbestfriend]

JROCKS, are these in conference? The teams shown before cutting off are out of conference.

Was writing this on the 1st post.

 

[JROCK1966]

JROCKS, are these in conference? The teams shown before cutting off are out of conference.

Was writing this on the 1st post.

aw shit my bad...I put in R for road when I was supposed to put in A for away...hold on.

 

[booksbestfriend]

Ex...UGA -2.5 away vs AU home

 

[JROCK1966]

JROCKS, are these in conference? The teams shown before cutting off are out of conference.

Was writing this on the 1st post.

here ya go...this is all conference games. I can't include them all because my screen cuts off but this one has your Ga/Aub game...

 

[JROCK1966]

Ex...UGA -2.5 away vs AU home

last two year Road conference favs have won 68% and covered 47.7% of the time....

I think the home scenario I first listed has better results....I have heard some teams get too comfortable at home and think they'll just be able to show up to win vs. inferior teams...

 

[booksbestfriend]

Thanks JR. Was hoping for a lower ATS percentage. Will say this tool is pretty quick. Amazing how fast you were able to get that.

 

[JROCK1966]

Thanks JR. Was hoping for a lower ATS percentage. Will say this tool is pretty quick. Amazing how fast you were able to get that.

there may be better results if I add some more conditions...like, for example, I could add that the road favorite has a better team defense to go along with what I've already done...the limitation for a scenario is really limited only by what the user can think of / come up with.....my head is hurting so I'm taking a break. I'll revisit sometime this week. Cheers!

 

[JROCK1966]

Thanks JR. Was hoping for a lower ATS percentage. Will say this tool is pretty quick. Amazing how fast you were able to get that.

here I did a quick one and it got you your lower ATS rate simply by adding the month of November....

 

[VirginiaCavs]

This stuff is so useful but all these numbers make my head spin while trying to digest them and organize them in my mind

 

[JROCK1966]

This stuff is so useful but all these numbers make my head spin while trying to digest them and organize them in my mind

Yeah, mine is jello right now.....but I now know I have a way to find hard evidence that will help me confirm or deny the existence of a situation...letdowns, bouncebacks, trap games, etc. etc....my mind is a mess right now because I have crammed the last week. Now that I have my code developed and saved it won't take long to use and I can settle down the cramming/crunching of the numbers. It's time well spent....now I need to go and exercise or something and get my mind off it...