can i pay someone to do this for me? this shits a pain in the ass. I think ive beat then line so far this season, maybe when im at work monday ill deal with setting these formulas up, but not tonight...
*I had a rogue line added in my previous list. I've removed that now but can't edit the post as I took longer than 5 mins messing around with the formatting*
Thanks Newc, understand median and mean well, however not sure I understand what CLV means. I am assuming it's the difference between what the bet line is as opposed to the closing line. I am guessing that a positive CLV would be good and indicate that the handicapper is choosing a side that is bet most often moving the line away from the number which was originally posted. However, would like confirmation of whether I am reading this correctly. Thanks,
CLV - close line value and it compares the line you bet vs the closing line. Consistently beating the closing line over the long haul should generate success.
Been messing with this for the past hour and I can't tie to the numbers above. Here is the formula Im using for first game listed above (-130 ---> -128)
I thought you wanted the percent change expressed per tribecalledjeff's methodology (ie as compared to the pinnacle close). You're using the method I originally suggested - a raw comparison between your line % value and pinnacle's close. There's nothing wrong with your calculation - there was -0.38% difference between your line and the close and this equates to -0.68% worth of the closing line.
I thought you wanted the percent change expressed per tribecalledjeff's methodology (ie as compared to the pinnacle close). You're using the method I originally suggested - a raw comparison between your line % value and pinnacle's close. There's nothing wrong with your calculation - there was -0.38% difference between your line and the close and this equates to -0.68% worth of the closing line.
OK, im sticking with this method since i figured it out finally, lol. Another question, can you breakout the math for when a dog flips to a favorite? For example:
This should help with your calcs - to save uploading spreadsheets and so on, if you want to automate your calculations, you could use the following input info in a spreadsheet. (The A1 through H1 cells are just headed text/expanations; copy the other cells' formulae as laid out then copy down your spreadsheet.)
A1 Coops line
B1 Pinnacle close line
C1 Coops line decimal
D1 Pinnacle line decimal
E1 Coops line percentage
F1 Pinnacle line percentage
G1 Raw line % difference
H1 Line difference expressed as percentage of close value
A2 your line (in US line notation ie -130, 150 or whatever)
B2 Pinnacle close line (as above)
C2 =IF(ISNUMBER(A2),IF(A2>0,SUM((A2/100)+1),SUM(100/A2*-1)+1),"")
D2 =IF(ISNUMBER(B2),IF(B2>0,SUM((B2/100)+1),SUM(100/B2*-1)+1),"")
E2 =IF(ISNUMBER(A2),SUM(100/C2),"")
F2 =IF(ISNUMBER(B2),SUM(100/D2),"")
G2 =IF(ISNUMBER(A2),(F2-E2)/F2,"")
H2 =IF(ISNUMBER(A2),(F2-E2)/100,"")
Interesting to see, I have a sample size of 28 games, I had the line close worse than me on 9 occasions. I am 8-1 when this happens. On the 19 games where I beat the close, I am 10-9. I know its such a small sample, but at least I chart this for the year now, where I will probably make close to 800 plays.
Looks like 11-1 and 7-9, but anyway straight-up record is of course of little relevance here. ROI does happen to tell the same general story (+9.11u, +33.3% roi and -1.69u, -4.6% roi) but over such a sample that's not all that meaningful either.
Anyway, I have your CLV thru yesterday (incl MLBx) at -0.64% roi and -0.42% eg.
Looks like 11-1 and 7-9, but anyway straight-up record is of course of little relevance here. ROI does happen to tell the same general story (+9.11u, +33.3% roi and -1.69u, -4.6% roi) but over such a sample that's not all that meaningful either.
Anyway, I have your CLV thru yesterday (incl MLBx) at -0.64% roi and -0.42% eg.
If you include MLBx it would be worse due to the mets play I posted @+150ish that closed around +240ish.
As I said in my main thread, at the end of the day I care more about units won then CLV, I'm just tracking it to see where I end up. I can almost say with certainty that my CLV was horrid last year but units won were great.
If you include MLBx it would be worse due to the mets play I posted @+150ish that closed around +240ish.
As I said in my main thread, at the end of the day I care more about units won then CLV, I'm just tracking it to see where I end up. I can almost say with certainty that my CLV was horrid last year but units won were great.
That's exactly why median is a better representation than mean. Outliers don't affect the median.
Thanks Jeff, and thanks to Clint its all automated now, just need to input the line I post and the Pinny close and the rest is done for me. I am learning more already, love how numerical the sport of baseball is.
Had another question relating to this. How would my weighting of plays come into play with CLV? Is there such thing as weighted CLV? For example my CLV on 5 Unit Angels play wa -1.82% (-163 --> -151). Would multiplying -1.82% by 5 (-9.0891%) be the proper way to weight CLV, if these is such a thing? Thanks again guys, keep schooling me, much appreciated!
Thanks Jeff, and thanks to Clint its all automated now, just need to input the line I post and the Pinny close and the rest is done for me. I am learning more already, love how numerical the sport of baseball is.
Had another question relating to this. How would my weighting of plays come into play with CLV? Is there such thing as weighted CLV? For example my CLV on 5 Unit Angels play wa -1.82% (-163 --> -151). Would multiplying -1.82% by 5 (-9.0891%) be the proper way to weight CLV, if these is such a thing? Thanks again guys, keep schooling me, much appreciated!
I'm not the expert the others are on this but it seems a 5u bet should count 5x, 4u 4x, etc so that you get an accurate estimate.
You definitely should not 'weight' your percent move record - it doesn't make sense at all (there is no sound mathematical reason for doing so). But it'd be useful for you to note separately where lines moved when your stake (and correlated +EV) was greater, or smaller. After all, when you're converting your line into a percentage, you're really only valuing your team's chance of success as a probability; your additional units should be laid out depending on how far off the book's line estimate is from your own in such instances. The amount of gain or loss in your percent worth per game should be enough to indicate if your perceived edge is smaller, or greater, than you originally thought. Your ROI should tell you all you need to know about the real worth of your additional units risked in any case - just keep two tabs of ROI and units risked, for weighted and unweighted games.
Lost % value where you have a greater amount risked (unless you have a tangible definite reason, like injury news or major personnel change after you lined the game) should signal that you're possibly accounting for something that isn't worth to the rest of the market what you believe it is, in your line. So it would be useful for you to see what typically happened to games with additional units staked just to see if you are perhaps over (or under) valuing some factor.
This is a good reason for analysing your lines in a percentage value - depending on your own thresholds for increasing your amount staked, it should be possible for a follower to play, at a reduced unit/stake, a side you have bet at multiple units, even if the line has moved in your favour, depending on how you value the worth of moving up in units staked.
And this is all assuming you do have a provable edge in these cases, of course.
To keep things honest for yourself, I would suggest the reverse should apply to your record-keeping too - if you get significant line help from some obvious factor that occurred after you lined the game you should note the probable reason for that move in your records as well as patting yourself on the back for scoring a nice line over the book
Good luck. Will be interesting to track. A lot more people complain about not getting their full expected value after beating the closing line in MLB than in NCAAB NBA NFL or NCAAF.
You definitely should not 'weight' your percent move record - it doesn't make sense at all (there is no sound mathematical reason for doing so).
So if a 1u bet beats close by 4% and another bet of 5u loses to close by 2% you would consider the aggregate result to be beating close? I wouldn't bet on your bankroll seeing it that way over time in deep markets.
I suggest interested parties direct their attention to finance where this is a well established concept.
So if a 1u bet beats close by 4% and another bet of 5u loses to close by 2% you would consider the aggregate result to be beating close? I wouldn't bet on your bankroll seeing it that way over time in deep markets.
I suggest interested parties direct their attention to finance where this is a well established concept.
Can that also get into determining the proper weighting to best maximize ones bankroll? (assuming you have shown an ability to recognize when your edge is bigger?)
So if a 1u bet beats close by 4% and another bet of 5u loses to close by 2% you would consider the aggregate result to be beating close? I wouldn't bet on your bankroll seeing it that way over time in deep markets.
I suggest interested parties direct their attention to finance where this is a well established concept.
Point taken re the aggregating of those weighted picks - to be fair, that's why I also wrote that the ROI should be assessed and individual noting of weighted lines should be undertaken as well. The OP wanted to calculate the difference in worth of their bet-at line and the closing line - using your example, a regimental 1u bettor would be delighted to follow along if the record was unweighted but the weighted variety would suggest not to follow.
Using the same example, the problem you cite lies squarely with money management (and likely calculation of EV) - the OP wanted to know the best method for tracking the difference between his lines and the closing equivalents and I don't think it should be weighted for this purpose.
Guys - i'm pulling my hair out here. How would i calculate CLV if i bet a team when at -104 and they closed +103? I can't get from Dog to Favorite, but some reason can't figure this out...
Comments
*I had a rogue line added in my previous list. I've removed that now but can't edit the post as I took longer than 5 mins messing around with the formatting*
-130 -128 -0.68%
-145 -140 -1.46%
-119 -114 -2.00%
136 122 5.93%
-145 -150 1.36%
115 119 -1.86%
-175 -181 1.21%
127 109 7.93%
118 114 1.83%
-116 -127 4.01%
136 140 -1.69%
-200 -210 1.59%
-116 -121 1.91%
-120 -106 -6.00%
145 136 3.67%
101 -116 7.36%
102 -104 2.89%
-185 -179 -1.18%
120 109 5.00%
Median % 1.59%
Mean % 1.57%
Think I fixed it for you. Not home and short on time.
Clint/Tribe/Goats
Been messing with this for the past hour and I can't tie to the numbers above. Here is the formula Im using for first game listed above (-130 ---> -128)
=(100/130)+1
1.7692
=100/1.7692
56.52
=(100/128)+1
1.78
=100/1.78
56.14
=56.52/100
56.5217%
=56.14/100
56.1403%
56.1403%-56.5217% = -0.38%
Clint came up with -0.68%
Any idea what Im doing wrong here?
OK, im sticking with this method since i figured it out finally, lol. Another question, can you breakout the math for when a dog flips to a favorite? For example:
+101 ---> -116
and when a fav flips to a dog:
-105 ---> +110
Thanks!
A1 Coops line
B1 Pinnacle close line
C1 Coops line decimal
D1 Pinnacle line decimal
E1 Coops line percentage
F1 Pinnacle line percentage
G1 Raw line % difference
H1 Line difference expressed as percentage of close value
A2 your line (in US line notation ie -130, 150 or whatever)
B2 Pinnacle close line (as above)
C2 =IF(ISNUMBER(A2),IF(A2>0,SUM((A2/100)+1),SUM(100/A2*-1)+1),"")
D2 =IF(ISNUMBER(B2),IF(B2>0,SUM((B2/100)+1),SUM(100/B2*-1)+1),"")
E2 =IF(ISNUMBER(A2),SUM(100/C2),"")
F2 =IF(ISNUMBER(B2),SUM(100/D2),"")
G2 =IF(ISNUMBER(A2),(F2-E2)/F2,"")
H2 =IF(ISNUMBER(A2),(F2-E2)/100,"")
MEDIAN: 0.83687%
MEAN: 0.63517%
OK now that we have that solved....help me with totals.
Anyway, I have your CLV thru yesterday (incl MLBx) at -0.64% roi and -0.42% eg.
If you include MLBx it would be worse due to the mets play I posted @+150ish that closed around +240ish.
As I said in my main thread, at the end of the day I care more about units won then CLV, I'm just tracking it to see where I end up. I can almost say with certainty that my CLV was horrid last year but units won were great.
Pinny CLV (No MLBX): 0.79% (Median) / 0.51% Mean
Raw CLV (With MLBX): 1.29% (Median) / 0.02% Mean
Pinny CLV (With MLBX): 0.79% (Median) / 0.22% Mean
That's exactly why median is a better representation than mean. Outliers don't affect the median.
Keep up the winners!
Had another question relating to this. How would my weighting of plays come into play with CLV? Is there such thing as weighted CLV? For example my CLV on 5 Unit Angels play wa -1.82% (-163 --> -151). Would multiplying -1.82% by 5 (-9.0891%) be the proper way to weight CLV, if these is such a thing? Thanks again guys, keep schooling me, much appreciated!
I'm not the expert the others are on this but it seems a 5u bet should count 5x, 4u 4x, etc so that you get an accurate estimate.
I agree, but Im not expert either, here are those results
Just going to include the 2 MLBX games going forward, much easier:
Unweighted Raw CLV: 1.29% (Median) / 0.02% Mean
Unweighted Pinny CLV: 0.79% (Median) / 0.22% Mean
Weighted Raw CLV: 1.87% (Median) / 0.91% Mean
Weighted Pinny CLV: 0.91% (Median) / 0.42% Mean
At least it looks like Im weighting properly according to the results.
(Unless the logic is flawed, lol)
Lost % value where you have a greater amount risked (unless you have a tangible definite reason, like injury news or major personnel change after you lined the game) should signal that you're possibly accounting for something that isn't worth to the rest of the market what you believe it is, in your line. So it would be useful for you to see what typically happened to games with additional units staked just to see if you are perhaps over (or under) valuing some factor.
This is a good reason for analysing your lines in a percentage value - depending on your own thresholds for increasing your amount staked, it should be possible for a follower to play, at a reduced unit/stake, a side you have bet at multiple units, even if the line has moved in your favour, depending on how you value the worth of moving up in units staked.
And this is all assuming you do have a provable edge in these cases, of course.
To keep things honest for yourself, I would suggest the reverse should apply to your record-keeping too - if you get significant line help from some obvious factor that occurred after you lined the game you should note the probable reason for that move in your records as well as patting yourself on the back for scoring a nice line over the book
I feel stupid in this thread...
Raw Mean: +0.02%
Pinny Median: +0.57%
Pinny Mean: +0.19%
+9.77 Units when line moves against me
+7.66 Units when line moved for me
44 Game Sample Size
Anything else I should be tracking?
I suggest interested parties direct their attention to finance where this is a well established concept.
Can that also get into determining the proper weighting to best maximize ones bankroll? (assuming you have shown an ability to recognize when your edge is bigger?)
Point taken re the aggregating of those weighted picks - to be fair, that's why I also wrote that the ROI should be assessed and individual noting of weighted lines should be undertaken as well. The OP wanted to calculate the difference in worth of their bet-at line and the closing line - using your example, a regimental 1u bettor would be delighted to follow along if the record was unweighted but the weighted variety would suggest not to follow.
Using the same example, the problem you cite lies squarely with money management (and likely calculation of EV) - the OP wanted to know the best method for tracking the difference between his lines and the closing equivalents and I don't think it should be weighted for this purpose.
So 49.26% - 50.98%= -1.72% CLV
For dogs, use 100/(100+odds), so 100/(100+103)
For faves, use odds/(100+odds), so 104/(100/104)