NASCAR FORUM

Comments

  • jt77316

    Daytona International Speedway –

    Schedule — all times are Eastern Time (ET)

    — Duel 1: Thursday, 7:00 PM (FS1)
    — Duel 2: Thursday, 8:45 PM (FS1)
    — Practice 1: Friday, 5:05 PM (FS1)
    — Practice 2: Saturday, 12:30 PM (FS1)
    — Race: Daytona 500, Sunday, 2:30 PM (FOX)

    Race Details

    — 40 Drivers Entered; 200 Laps, 500 miles
    — Stage 1: Lap 1 – 60; Stage 2: Lap 61 – 120; Stage 3: Lap 121 – 200
    — Expected Fuel/Pit Window – xx-xx Laps

    Note: Fuel Cell capacity is an estimate.

    Daytona International Speedway –

    2.5 mile Super Speedway

    Similar Tracks: Talledega

    Last Five (5) Race Winners

    — 07/01/2017: Ricky Stenhouse, Jr.
    — 02/18/2018: Austin Dillon
    — 07/07/2018: Erik Jones
    — 02/17/2019: Denny Hamlin
    — 07/07/2019: Justin Haley

    Last Five (5) DK and FD Fantasy Points Leaders

    07/01/2017: Brendan Gaughan (66.00 DK FPts) – Paul Menard (64.8 FD FPts)
    02/18/2018: Ryan Blaney (65.00 DK FPts) – Austin Dillon (70.3 FD FPts)
    07/07/2018: Erik Jones (78.75 DK FPts) – Erik Jones (73.9 FD FPts)
    02/17/2019: Kyle Busch (82.25 DK FPts) – Kyle Busch (78.9 FD FPts)
    07/07/2019: Justin Haley (79.75 DK FPts) – Justin Haley (72.3 FD FPts)

    Data from Last 5 Daytona Races

    Driver Rating:

    1. Ricky Stenhouse, Jr. – 95.9
    2. Alex Bowman – 91.8
    3. Ryan Blaney – 90.7
    4. Kevin Harvick – 87.9
    5. Jimmie Johnson – 85.3

    Laps Led:

    1. Ryan Blaney – 141
    2. Ricky Stenhouse, Jr. – 97
    3. William Byron – 56
    4. Joey Logano – 54
    5. Kevin Harvick– 27

    Fastest Laps:

    1. Ty Dillon – 31
    2. Jimmie Johnson – 26
    3. Michael McDowell – 25
    4. Ricky Stenhouse, Jr – 19
    5. Clint Bowyer – 16

    Race Data, News, and Information

    Rotogrinders NASCAR Free tools
    Motor Racing Network
    Racing Reference

    Race Day Breaking News and Weather

    Stephen Youngstevietpfl
    Bob Pockrass
    NASCAR Weather
    Jayski

    Note: As needed information above will be updated

  • marker0357

    I’m thinking of a dart on Tyler Reddick ahead of the Duels. Right now, it looks like he is 40:1. Your input is appreciated.

  • gvn2fly1421

    Any ideas on how FanDuel is planning on doing the Duels? They have one contest out there for both races? Or is the contest only for the 1st race?

  • gje627

    FYI If anybody’s interested in my numbers for the Daytona 500, I won’t be posting here … too difficult with all the formatting (early in the week, and I already have a little north of 30 tables and charts) … but I will have them.

    As of right now, I am either finished or mostly finished with DK, FD, and finish position race factor correlations (everyone seems to love the correlations), DK and FD projected fantasy points needed to win, both Cash and GPP (including breakdowns by scoring category for each site), all DK and FD historical optimal lineups, including breakdowns by scoring category), in-race outcome factor analysis (cautions, lead changes, passing) and finishing factor analysis (lead lap, DNF crash, DNF mechanical)

    Later this week I’ll have probability tables, projected fantasy points for each driver, probability simulation results, and … a few more things. Everything will include both DK and FD, and finish position for those who bet but don’t play fantasy.

    Anyways, if anyone is interested in any of the analyses I’m doing this week, just let me know by replying below. Pretty busy the remainder of the day, so I may not be able to reply until later tomorrow.

  • DefinitelyMiami

    I’m interested

  • gje627

    @DefinitelyMiami said...

    I’m interested

    Definitely Miami !!!

    Get it ??? !!! :)

    Dang … I crack myself up !!!

    I’ll be in touch tomorrow with a download. :)

  • gje627

    Okay ….

    Said I wouldn’t do it; but Hey, one table never hurt anyone, right?

    Nonetheless, below are “PairWise” rankings for each driver with a minimum of 30 starts last year.

    You may have seen PairWise rankings in other sports, particularly college sports (especially college hockey … my absolute favorite sport !!!). It is based in Ranking Theory (shocking !!!) using mostly discrete mathematical methods, where each team/individual (in our case, driver), competes head-to-head, across a number of variables, against every other driver and weighted points are earned by a driver for beating an opponent in each category, where win = 1, and lose = 0 (discrete, again these are then weighted per variable), under consideration.

    The weights I incorporated in my model are altogether based on the correlation of certain variables in all of the three categories below. For simplicity, I used Spearman’s Correlation for ranked data, as opposed to the more common Pearson’s correlation to determine weights. Since Spearman, unlike Pearson, does not measure linearity and the magnitude of difference between variables is not continuous, the categorical “points” difference between drivers will naturally be less than had the analysis been performed using Pearson. Nevertheless, while correlation is considered, causation is not. Primarily because for purposes of PairWise, identifying the exact nature of the relationship is not necessary.

    Other than variable weights, other weights were applied, most importantly track and track type. For the DK and FD fantasy categories, a noticeable weighting omission was a salary for each driver.

    Honestly, the reason I did not incorporate salary … was … well … for the life of me, I couldn’t figure out a statistically/mathematically valid method to do so. Suffice to say … salary should be incorporated given that salary cap is the singular lineup constraint in NASCAR on both DK and FD.

    Anyways, if you’re still interested, see below ….

    If, not, I won’t hold a grudge if you report me to the mods for wasting both space … and your time. :(

    2019 NASCAR Cup – Final Driver PairWise Points and Rankings

    —-

    Name DK Points FD Points Finish Points DK Rank FD Rank Finish Rank
    ————————————- ——————— ——————- ———————— ——————- ——————- ————————
    Kyle Busch 57.6 57.6 57.6 1 1 1
    Martin Truex 55.8 55.8 54 2 2 3
    Kevin Harvick 54 52.2 52.2 3 4 4
    Denny Hamlin 52.2 54 55.8 4 3 2
    Joey Logano 50.4 50.4 50.4 5 5 5
    Brad Keselowski 48.6 48.6 48.6 6 6 6
    Ryan Newman 46.8 46.8 41.4 7 7 10
    Ryan Blaney 45 43.2 45 8 9 8
    Kurt Busch 43.2 45 46.8 9 8 7
    Kyle Larson 41.4 39.6 37.8 10 11 12
    Chase Elliott 39.6 37.8 36 11 12 13
    Alex Bowman 37.8 41.4 43.2 12 10 9
    Erik Jones 36 36 32.4 13 13 15
    William Byron 34.2 32.4 39.6 14 15 11
    Chris Buescher 32.4 28.8 23.4 15 17 20
    Clint Bowyer 30.6 34.2 34.2 16 14 14
    Daniel Suarez 28.8 30.6 28.8 17 16 17
    Matt DiBenedetto 27 23.4 21.6 18 20 21
    Paul Menard 25.2 27 27 19 18 18
    Aric Almirola 23.4 25.2 30.6 20 19 16
    Ty Dillon 21.6 19.8 16.2 21 22 24
    Jimmie Johnson 19.8 21.6 25.2 22 21 19
    Austin Dillon 18 18 19.8 23 23 22
    Bubba Wallace 16.2 12.6 10.8 24 26 27
    Ryan Preece 14.4 14.4 12.6 25 25 26
    Ross Chastain 12.6 3.6 1.8 26 30 31
    Ricky Stenhouse 10.8 16.2 18 27 24 23
    Daniel Hemric 7.2 10.8 14.4 28 27 25
    Corey LaJoie 5.4 7.2 7.2 29 28 29
    Michael McDowell 3.6 5.4 9 30 29 28
    Landon Cassill 1.8 0 0 31 32 32
    David Ragan 0 1.8 3.6 32 31 30
    Matt Tifft 0 0 0 32 32 32

    —-
    END

  • jt77316

    @gje627 said...

    FYI If anybody’s interested in my numbers for the Daytona 500, I won’t be posting here … too difficult with all the formatting (early in the week, and I already have a little north of 30 tables and charts) … but I will have them.

    As of right now, I am either finished or mostly finished with DK, FD, and finish position race factor correlations (everyone seems to love the correlations), DK and FD projected fantasy points needed to win, both Cash and GPP (including breakdowns by scoring category for each site), all DK and FD historical optimal lineups, including breakdowns by scoring category), in-race outcome factor analysis (cautions, lead changes, passing) and finishing factor analysis (lead lap, DNF crash, DNF mechanical)

    Later this week I’ll have probability tables, projected fantasy points for each driver, probability simulation results, and … a few more things. Everything will include both DK and FD, and finish position for those who bet but don’t play fantasy.

    Anyways, if anyone is interested in any of the analyses I’m doing this week, just let me know by replying below. Pretty busy the remainder of the day, so I may not be able to reply until later tomorrow.

    I’m always interested.

  • gje627

    @jt77316 said...

    I’m always interested.

    Got it …. I’ll make sure you get everything. Later tonight, as opposed to later today, is quickly becoming a better time estimate as I’m working on some other things and therefore won’t be able to get back to NASCAR until sometime this evening.

  • jbles311

    im interested

  • gje627

    @jbles311 said...

    im interested

    Okie-Dokie ….

  • gje627

    Sorry guys …. I said tonight on some numbers, but no-go.

    Soon … I promise.

  • Stangs13

    Any thoughts on the duels? These seem like the only superspeedway races that we don’t want back markers. Duel 1 I like Penske. Trying to decide which cheapie to put in out of Haley Sorensen Nemechek and Suárez.
    Duel 2 I like Ganassi, and the two Gibbs guys but not sure who to put with them. I feel like the expensive drivers in that race aren’t as good values.

  • gje627

    In partial response to Stangs13 … some fun facts and food for thought (but no real recommendations) if you’re playing tonight’s duels.

    DK Duel 1 — 22 Drivers (select 6): 70,453 possible lineups at or under the $50K salary cap

    DK Duel 2 — 21 Drivers (select 6): 51,543 possible lineups at or under the $50K salary cap

    FD Combined — 43 Drivers (select 5): 945,736 possible lineups at or under the $50K salary cap

    — Ricky Stenhouse, Jr and Alex Bowman start on the pole in each respective duel. Both are likely unplayable based on their starting position and the fact they are the only two drivers who have already earned starting positions for the Daytona 500. (1. Stenhouse, Jr – front row, pole; 2. Alex Bowman front row in lane not chosen by Stenhouse.) so have very little reason to be aggressive tonight (though, see below regarding points). Side Note: There is no “choice” for rows 2 through 20. Duel 1 will set the inside row, and Duel 2 the outside row.

    — There are 43 cars in both duels combined, but per NASCAR rules a maximum of 40 drivers can race in the Daytona 500 (or any other regular season/non-exhibition race) . Of the 43 drivers, 36 charter team drivers are guaranteed a spot in the Daytona 500; meaning 3 open drivers tonight will not make the Daytona 500. … Not to brag, but I did all that math in my head while typing …. Anyone impressed. :)

    These drivers are:

    Duel 1: Daniel Suarez, Justin Haley, Reed Sorenson, Chad Finchum (top 2 drivers will make the 500)
    Duel 2: Brendan Gaughan, JJ Yeley, Timmy Hill (top 2 drivers will make the 500)

    The significance? With 7 drivers for 4 spots, expect these drivers to focus more on competing against one another, instead of for the win.

    Note: More “head” math. :)

    — Expect fewer cautions tonight than occurred in Sunday’s Busch Clash race since, unlike the Clash, this is a points race. Drivers tonight finishing Top 10 in each race earn points in a manner equivalent to points awarded for Top 10 stage finishes during the regular season (10 pts. – 1st, 9 pts. – 2nd … 1 pt. 10th). Thus, drivers will (may?) be more likely to settle for a handful of points instead pushing for a win (even while points awarded are minimal, we all know every point counts when it comes to the playoff cut line). In addition, remember the Clash was all about the prestige of winning (and of course the money). Really, the duels are only about qualifying position for the 500 (and making the race for open cars), so again, the motivation to actually win each race is pretty low, especially since starting position matters very little at Daytona.

    ——————-

    On a different note, I owe some people here some numbers, and I will be uploading a small handful of stuff to my Google Drive in the next “hour-ish”. I will send those this message pertains to a download link once finished. On a side note, the reason this is taking longer than it should is because I am in the process of moving all my databases and analysis online, so, with some files on my hard drive, some in Google Drive, some in my DropBox account, and some now in my online MySQL database, I’m struggling a little bit to keep organized and find everything.

  • gje627

    Okay, for those who were waiting for some NASCAR/Daytona 500 numbers … I PM’d each of you a Google Drive link to an Excel file. If for some reason, you did not receive and still want the file, please reply below.

  • jt77316

    @gje627 said...

    Okay, for those who were waiting for some NASCAR/Daytona 500 numbers … I PM’d each of you a Google Drive link to an Excel file. If for some reason, you did not receive and still want the file, please reply below.

    Very nice. It’s much appreciated. Now to see if I can make sense of it all, or any of it!

  • gje627

    Yeah, sorry about that. I did provide a brief explanatory section on the first tab. Not great, but it is something. An overall comment about one small thing I included in the correlations-only … the “p-value”. This is just a general statement, but when someone provides correlations, never-ever trust them unless they also report a “p-value” or at least indicate that all p-values are < .001 like I did …. If you notice the DK and FD optimal p-values, there are several that are relatively high. This mainly has to do with sample sizes being reduced by filtering to a sub-population (only includes drivers in optimal/standardized optimal lineups and not all drivers in the field). Per the table, these are not really too desirable, but it does it not mean the correlations are not usable, it just means that you are taking more risk in using them.

    Regarding the confidence intervals, for the most part that’s just extra … but very helpful. The rule of thumb there is when using two-sided confidence levels the narrower the range between the two numbers … the better.

    Lastly, regarding the projected points numbers. There have only been 30 NASCAR races at Daytona since 2005, 3 FD contest races, and 9 DK contest races. Thus, the samples there are really small.

    Thus, I only used those races to build a bigger data set. In my case, I did this by performing Monte Carlo simulations (2,000 iterations (runs)).

    That is what the n=2000 number means in the first row of those tables.

    The Monte Carlo numbers are statistically significant and valid.

    Should have mentioned these things in my PM.

    Anyways, if you have questions, please don’t hesitate to ask. My only request, is that you ask in the forum and not via PM; that way if anyone else has the same question they will see the answer as well

    P.S.: Regarding that last statement, the only reason I PM’d the file and did not post here was because I hate posting direct links to my accounts in a public forum … and for better or worse … I trust everyone I sent the link to. :)

  • jt77316

    I am curious about one thing and maybe it is easy to explain. On the DK optimal only correlation sheet, starting position is one of the higher Spearman correlations (.61). Considering how the finish correlates to FD points (.72), I completely understand, but back to the starting position, what influences that to make it so high. It seems we want to fade those at the front because of the likelihood of wrecks and negative place differential, but if the starting position correlates that well with finishing, what am I missing?

  • gje627

    @jt77316 said...

    I am curious about one thing and maybe it is easy to explain. On the DK optimal only correlation sheet, starting position is one of the higher Spearman correlations (.61). Considering how the finish correlates to FD points (.72), I completely understand, but back to the starting position, what influences that to make it so high. It seems we want to fade those at the front because of the likelihood of wrecks and negative place differential, but if the starting position correlates that well with finishing, what am I missing?

    Great Question !!!

    One thing about correlations is that you kind of have to go on an acid trip sometimes to understand.

    See if I can explain (briefly … ouch).

    The correlation here does not mean:

    Start 1 Finish 1 in DK FPts
    Start 2 Finish 2 in DK FPts
    etc.

    It simply means, only for those drivers finishing in the optimal/standard optimal lineups the two numbers move together. Remember, here (for DK) we are eliminating 34 drivers from consideration and only looking at the top 6 (provided they are within salary constraints).

    The positive correlation indicates, that for those strictly “winning lineup” drivers only:

    1) For the most part they move together (regardless of starting position), and
    2) They move in the same direction (when one moves up the others move up as well).

    Again,, not considering the whole field, just drivers in the winning lineup (very, very, important here) this could mean, and for the Daytona 500 likely does mean something more like this (an example only):

    Start 34 Finish 1 in DK FPts
    Start 37 Finish 2 in DK FPts
    Start 29 Finish 3 in DK FPts

    Here the drivers are moving “mostly” together, in “very much” the same direction but not at a 1.0 Spearman, but still they are moving mostly together.

    Also remember, correlation measures direction and strength, but nothing else (Spearman does not measure linearity / Pearson does, fwiw).

    To further evaluate this “relationship” I would move on to Regression Analysis, using the actual continuous variable, as opposed to the discrete. I.e., simple linear regression as opposed to logit/probit or specialized (complicated !!) regression. Here, Pearson and not Spearman will be reported.

    Assuming someone here wants to discuss normal distributions … binomial distributions, poisson distributions … I get your point, but can we make that a conversation for later?

    Does this make more sense?

    Or is it … OUCH !!!

    EDIT: The very simple explanation (should have started with this) …. “Selecting the correct drivers based on starting position will likely be very important (but not absolutely critical … all the drivers starting in back could get caught-up in the “big one” wreck) in determining who will likely be in the DK winning GPP lineup on Sunday.”

  • starstx

    Glad to see you back on the forum gje And enjoy reading all of your stuff. Even if I only understand a fraction of it lol
    … I first started playing Nascar dfs year before last and you were killing it. GL again this year.

    I have a non math non data set question but it does involve numbers technically. I play DK and most of the sports which are all a little different. In PGA dfs I am comfortable leaving up to $500 salary left over and every once in awhile dip lower. But with Nascar it seems I should loosen that up quite a bit and be comfortable leaving much more on the table sometimes. I was just curious how much salary are you comfortable leaving on the table and do you have a limit you will not go under?

    This question is for most normal Nascar races that typically have a handful of drivers not finish. For Daytona … I got no constraints!

  • KingzLanding813

    I’m interested…thanks in advance

  • elmore0631

    Any drivers gonna be looking or having a reason to win duel

  • gje627

    @starstx said...

    Glad to see you back on the forum gje And enjoy reading all of your stuff. Even if I only understand a fraction of it lol
    … I first started playing Nascar dfs year before last and you were killing it. GL again this year.

    I have a non math non data set question but it does involve numbers technically. I play DK and most of the sports which are all a little different. In PGA dfs I am comfortable leaving up to $500 salary left over and every once in awhile dip lower. But with Nascar it seems I should loosen that up quite a bit and be comfortable leaving much more on the table sometimes. I was just curious how much salary are you comfortable leaving on the table and do you have a limit you will not go under?

    This question is for most normal Nascar races that typically have a handful of drivers not finish. For Daytona … I got no constraints!

    Hey starstx … great to hear from you again !!!

    Regarding salaries, particularly in NASCAR, they tend not to mean a whole lot with the obvious broad principle that better drivers are priced higher than worse drivers … after all, salaries are set simply by some analyst crunching the numbers … just like you or I can (note: DK NASCAR analysts … excellent; FD NASCAR analysts … okay, but mostly … WTF?)

    Interestingly, there is a “secret sauce” formula you can use to determine how much salary you can be comfortable leaving on the table given a particular, specific race (warning … caution … warning … “particular, specific race” in bold … important, important, important).

    Anyways, even though the sauce is secret … I can explain in a later post, if that’s cool. Also, even though I said I wouldn’t, I may post an “illustrative” table to demonstrate …. need to grab a beer and eat some food now … haven’t had a speck of food since yesterday morning so probably later tonight.

    Again, thanks for stopping in and saying “Howdy !!!” :) …. I’ll post again in a little bit.

  • gje627

    Real quick … @jt77316’s excellent question, got me thinking … I don’t think my sub-set correlations (DK/FD optimal) measure-up … they’re fine, in a way, likely not that useful, though.

    I think I need a different approach for sub-sets, particularly where one variables must be ranked (or even generally ordinal) …. If I can find my z-scores in my toolbox I may be close … if I have to revert to log transformations … somebody please call 911. Just say gje627 in Colorado and they will know what you’re talking about !!! HA-HA … LOL. :)

    Anyways, in the meantime, for those who got my “file”, please take with a grain of salt my DK/FD optimal correlations. Regarding the “all” numbers, they should be fine and good as they can get.

    Lastly, my long-winded reply to @jt77316 still stands. The explanation is correct, just worried the numbers in the categories I list may actually not be that helpful, is all.

  • stevengoo

    Is it too late to say I’m interested??

  • gje627

    Good Luck to the “Duelers” …. I don’t play exhibition races anymore, but I’ll be rooting for you !!!

    Also, FYI, only play small in “drafting” races (formerly “restrictor plate” races … Daytona/Talladega … the last plate race was last year’s Daytona 500, now every race has large or small tapered spacers).

  • X Unread Thread
  • X Thread with New Replies*
  • *Jumps to your first unread reply

Subforum Index

RotoGrinders.com is the home of the daily fantasy sports community. Our content, rankings, member blogs, promotions and forum discussion all cater to the players that like to create a new fantasy team every day of the week.

Bet with your head, not over it!
Gambling Problem? Call 1-800-Gambler