[avssakic]

Yea sometimes history and trends pay off more than qual and prac speeds

[onlooker]

I mix all that together. It did great last year, but this year hasn't been so good.

[Optional]

Do any of you guys bother with statistical handicapping for NASCAR?

[hawk 5]

Do any of you guys bother with statistical handicapping for NASCAR?

I take historical stats to grab early drivers. But I also consider who's hot and who's not.

[Optional]

Pre-practice, I take the last 2 races, last 2 races at this track, average at this track compared to career, championship position, cumulative season driver rating and the opening lines.

I rank the drivers in each one (without concern for the actual numbers) and then weight each metric to come up with a number that I convert to a line.

Be interested to hear if anyone can see an angle I might be missing there, or knows of some other stat they think I should/could take into account, pre-race.

[dew22]

Do any of you guys bother with statistical handicapping for NASCAR?

Similar to your pre-practice layout Optional, but then I formulate figures from Happy Hour between fastest lap, average speeds in HH, and a few other things. I've tracked it back 3 seasons, and it predicts top-10 finishes at almost a 70% clip and top-5 at 40%--not too hot at predicting winners though. Works really well at some tracks, but so-so at others (short tracks mainly). These statistics are probably irrelevant in the long run, but they help confirm my gut feelings and have helped me make a good bit of coin in the past. There are some great bargains in the books!

Oh, and as someone else said, recent form is a must in capping. I worked in Cup for a few years about a decade ago, and when a team is on a good run, they've generally hit on something noteworthy in gaining extra horsepower, or a chassis balance on a specific car that translates to success almost anywhere they run. How long they ride the momentum is a different story.

Some drivers I automatically toss, and I'm guessing everyone else does as well. Some just suck, and others are in cars that scream mechanical failure. I don't care how well they practiced or qualified. Some drivers and some cars perform great at certain tracks too. For this weekend the #2 should do well, barring any on-track incidents. Over the last twenty years, you could put anyone in the car and they'd compete. Penske used to run a unique shock package that worked wonders there. I don't know if they still use the same setup, but I'm sure it's a fall back if things are going sour in practice.

Bristol is a tough one to predict, but sure is fun as hell to watch. I'd be happy if they raced there ten times a year!

[Optional]

Thanks Dew, that's interesting stuff. I only started developing my model this year, and have really only got to the point I feel it's nearing perfected in the last half dozen races or so. Because I use numbers that there is no way to get historically, I can't really back test it, but I have had a run of 4 race winners in a row, and several weeks with the top 5 picks all finishing top 6 or 8. (It's been up and down for the last few races though)

I'll take extra care with the short track races. Obviously I haven't had a chance to work out stuff like that myself as yet, so thanks.

I only mentioned pre-practice stuff, as I assume we all use the practice and qually times pretty much the same way. Although I give 60% of the weight to Top Speed and 40% to the average time in each session, which might be a little different. I wish we could just get the raw loop data, that would open up a bunch of useful angles to check.

I toss out most drivers from my final calcs like you too. I do a calculation to work out each driver's probability of beating any other randomly chosen driver in a head to head matchup, then anyone with less than 50% gets dumped. That usually leaves ~16 drivers I'm interested in.

I hope my numbers pick up anyone on a hot form streak, or with an extra special liking for the track, and avoid adding any gut feel to the equation.

I feel like I know enough to be able to pick winners based on gut feel, reading news and common sense, but the evidence of last year, backing only 4 or 5 winners in 36 races, doesn't support that feeling. And I'm at nearly 300% of my starting (small) bankroll so far this year, so my gut no longer gets any say!

Although come to think of it, changing my metric weightings for each track does bring in a gut feel element.

I agree there seems to be bargains betting NASCAR I never see elsewhere. I started out modeling with F1 (which I do well with) and that market is so efficient there is almost never a winning chance at generous odds. Although if i had a choice I'd prefer NASCAR to be that liquid and tight too, as I love scalping the market on BetFair. Turns each race into a week long betting game where I end up making heaps of plays.


Do you think the Penske magic applies to Kesolowski's chances at Bristol too? The way he can dominate in Nationwide I have been expecting him to turn into a weekly top 10 chance for ages, yet he keeps looking out of his depth week after week.

[Optional]

so my gut no longer gets any say!

I just realized that is a lie. Just last week I layed against Jimmie, even though my model said he should win. (he was just way too short, so it +EV anyway, but my gut feeling about his history of bad 'luck' at the track swayed me)

That was very unusual for me this year though.

[dew22]

I wish we could just get the raw loop data, that would open up a bunch of useful angles to check.

This would be wonderful

I'm at nearly 300% of my starting (small) bankroll so far this year, so my gut no longer gets any say!

Don't change a thing!

Do you think the Penske magic applies to Kesolowski's chances at Bristol too? The way he can dominate in Nationwide I have been expecting him to turn into a weekly top 10 chance for ages, yet he keeps looking out of his depth week after week.

We'll see how he practices, but I'm thinking there will be some value in a top ten finish. I can't remember the first race this season, but he did lead a handful of laps. He'll need to put in a better qualifying effort to win. Isn't it something like 80% winners here start from the top-10? As we've seen year after year, most drivers lose their cool trying to work through the traffic here and hurt their chances at a good finish.

Just last week I layed against Jimmie, even though my model said he should win. (he was just way too short, so it +EV anyway, but my gut feeling about his history of bad 'luck' at the track swayed me)

Good call. Why didn't you step in and caution me on Gordon?

[Optional]

I was with you there. I thought Gordon was the best chance he has been for ages! Although the qualifying position did put me off a tad

[dew22]

I give 60% of the weight to Top Speed and 40% to the average time in each session,

I usually give average speed an edge over top speed. This is probably why my model stinks at picking winners. One factor that causes me to shift this balance is if Goodyear is using a new tire at the track from the previous race. The harder the compound, the more edge I give to top speed, since the tire may last longer throughout the race. If this is the case, I also try to watch Happy Hour and see who is getting off the corners best, since the harder compounds make the car slippery and difficult to bite into the track on acceleration.

Another angle I've made some money on is betting on qualifying head-to-head, although I don't play this very often. History generally helps a ton here, but I also look to see what chassis the teams have brought to the track. If they've won poles or qualified very well in the car before, bingo.

Note: Jayski will often list what chassis some teams have brought to the track

[Optional]

I go with top speed more simply because I don't trust the average figures. In fact I left them out totally originally as I thought they would definitely introduce at least some errors, and potentially large ones.

It became obvious, comparing results with and without them, that they were important though.

Where do you bet the quali head 2 heads? I've been thinking about opening a Bwin account as they are the only place I've seen offering them.

Nice tip about hard tyre compounds too. I'm starting to feel like a newbie all over again.

[dew22]

5D. Great options. When you get an account PM me and I'll give you my thoughts on line movement there.

Yeah, I've had similar doubts on the average figures, and I've flip-flopped throughout the years on how much I weigh them in capping. With this weekend I'll definitely lean toward top speed. The concrete is too nice on tire wear

[Optional]

Thanks Dew, you're a good bloke.

I plan get more serious with my punting next year, and have been putting off opening accounts at a couple of books I think I'll use, like 5Ds. So I can maximize my bonus. Figured I will get the most making a decent deposit when I first open it.

Might PM you anyway.

[rickie65]

what I usually do, time permitting....my info primarily comes from MotorRacingNetwork (used to be RacingOne). I use the last 5 years link to get finishing positions for the last 5 years, put them in excel and color code top 5 finishes and a 2nd color for 6th-15th finishes so they stand out. I add in the starting lineup and Happy Hour times (color-coded again, of course) and just look for consistency. I also add in the COT finishes only which is usually last 3 years looking for an edge there too. And at tracks like Bristol I pay special attention to just the 2nd races of the previous years because night and day results can be as different as, well...night and day. the MRN site also has chassis info on many drivers but last time I used it I was way off in my picks. And I keep recent finishes in the back of my mind too so I don't forget to consider momentum streaks good and bad.

[Optional]

The harder the compound, the more edge I give to top speed, since the tire may last longer throughout the race. If this is the case, I also try to watch Happy Hour and see who is getting off the corners best, since the harder compounds make the car slippery and difficult to bite into the track on acceleration.

What do we do now that none of the cars appear to able to get off?

[dew22]

What do we do now that none of the cars appear to able to get off?

You watching this? A lot of teams are struggling right now. Johnson and Gordon dialed in pretty nice during the 1st practice. Newman is looking like a solid top ten finish.

I picked Montoya over Kurt Busch and Edwards over Kenseth for Qualifying. Read Montoya tire tested the new package...I'm guessing he'll take Busch in Q

[dew22]

Hamlin's car looks great rolling through the corners. Smooth sailing, and sucked down to the bottom of the track.

[Optional]

I thought Montoya and McMurray both looked like they were happy and better dialed in than most from the word go too.

I don't like betting on the JGR cars until I see them back on top of their game, but Denny is making me nervous with that decision so far, and your assessment of his quali isn't helping.

I might go out on a limb and take a few long shots this week. Keeping a close eye on Mark Martin.

[dew22]

You watching this? A lot of teams are struggling right now. Johnson and Gordon dialed in pretty nice during the 1st practice. Newman is looking like a solid top ten finish.

I picked Montoya over Kurt Busch and Edwards over Kenseth for Qualifying. Read Montoya tire tested the new package...I'm guessing he'll take Busch in Q

[Optional]

Congrats Dew.

Lets hope that's the start of a weekend long trend around here.

[merk1986]

ricky boby #1 woooohooooo