Quick Tips:
- A comma separates legs
- A "/" symbol separates sides of a market
- The first number in a leg is the side of the market you are wagering on.
- If a leg has multiple sides in its market (ie: superbowl winner), then separate all sides by a "/" symbol like this: +500/+250/+2000
SGP correlation Tips:
If you have a SGP that is for DeRozan to score 30+ points and DeRozan's team, the Bulls, to win the game, there is a positive correlation and SGP odds reflect that. You can use the devigger to place all combinations of a SGP to estimate fair odds with correlation like:
1. DeRozan 30+ and Bulls to cover = +180
2. DeRozan 30+ and Heat to cover = +250
3. DeRozan <30 and Bulls to cover = +390
4. DeRozan <30 and Heat to cover = +260
Plug that into the leg odds box like so: 180/250/390/260
It will output the fair odds of DeRozan 30+ points and Bulls to cover, taking into account the positive correlation between the 2 legs.
Juice Tips:
- If you only have one side of a market, you can specify the juice amount and the app will calculate fair odds based on the juice you specify.
To Specify juice, simply format like so: +285/15%
- If you have a similar market from that sportsbook that you'd like to derive the juice from and use that market's juice,
To Derive juice from another market, simply format like so: +285/[-116/-106]
Mutually Exclusive Events (XOR):
- If you have a bet on a single event, like a Nascar race, and the bet is for either Player X OR Player Y OR Player Z to win, you can format like so:
+500^+800^+1000/1200/1600/2000
In this example, there are 6 total race car drivers and you need X, Y, or Z to win. Notice that the "^" symbol combines those sides of the market so that only one needs to win. Other sides of the market should be separted by a "/" as normal. Note, this is mutually exclusive because there can be only one winner, not two.
Independent Events (OR):
- If you have a bet on multiple events and you only need one of them to hit, such as either Player X to get 25+ points OR Player Y to get 10+ rebounds, you can format like so:
-115/-110||-185/+140
Notice that instead of a ",", we used a "||" to establish that either of these 2 events can hit in order to win our bet. You can, of course, add other legs that are also "||" and you can also add other legs that are "," to specify that they must hit, such as Player X to score 25+ points OR Player Y to get 10+ rebounds, AND their team to win the game.
Events separated by a "||" are independent events and both could win, both could lose, or either could win/lose. As long as one of them wins, then the combined legs will count as a win.
Fair value:
- If you already have the fair value odds for something, for example you know something is fair value of +200, you can simply put +200 without any "/" symbol(s). It will assume the other side of the market is -200.
Correlation:
- If you can't do every combination of a SGP to derive correlated fair odds, you can use the correlation textbox.
- If you already know the correlation coefficient that you like'd to apply to your parlay, simply type it in like, "0.2"
a value of 1 would be 100% positive correlation (if one hits, the other will hit). A value of -1 would be 100% negative correlation (if one hits, the other can't hit).
- If you don't already know the correlation coeficient, you can derive it by supply the SGP odds and that sportsbook's one-sided odds for each leg.
For example, let's say we are doing a SGP for Chiefs to win and Travis Kelce to score a touchdown.
FanDuel has only one side for Travis to score a touchdown and we don't have 2 sides to the market.
We know that Chiefs are a -210 on FD and Travis Kelce is a -135 to score a touchdown.
Their SGP together is +134
We can provide the correlation box with:
+134=-210,-135
It will then derive the correlation coefficient from the sportsbook and apply it to the devig's fair odds.
If you were to bet 2 uncorrelated odds that were -210 and -135, the parlay would be +157.
The SGP above at +134 is lower (and lower payout) because Chiefs and Travis Kelce have positive correlation.
Since they are positively correlated, it is more likely to hit and thus the odds/payout is lower.
You can use FD or DK's SGP to derive correlation and then use a sharp book like Pinny in the Leg Odds box as the actual odds to devig.
Boosts:
If you have a profit boost on a site and you are trying to find the best bet to use it on.
Your Final Odds can be your unboosted odds and you can simply use the "Boost" checkbox to apply your boost percentage.
If it is a profit boost, choose "Profit Boost". If it is boosting the entire wager (such as BetMGM sometimes does), then choose "All".
This saves you time from having to continue to adjust the Final Odds textbox and adjusting it to be the boosted odds.
Multiplicative/Normalization/Traditional Devig Method:
- This is the normal way that we devig, but it does not take into account underdog bias. This method has a +900/-3000 outputting +968 fair odds.
Additive Devig Method:
- This is another way to devig. It helps to take into account underdog bias. This translates to a +900/-3000 outputting +1412 fair odds.
Note that there are rare instances where the fair value will be a negative percentage. This is normal and expected of the Additive devig method.
Power Devig Method:
- This is another way to devig. It exponentially takes underdog bias into account. This translates to a +900/-3000 outputting +2192 fair odds.
This method heavily leans towards heavy favorites.
Shin Devig Method:
- This is another way to devig. For Shin, 2 way markets will devig to the exact same fair odds as Additive. However, when devigging 3+ way markets, Shin will produce different fair odds than Additive.
Worst-case Devig Method:
- This method will always use the lowest implied probability of either Multiplicative or Additive or Power or Shin (set in the settings). This way you play it safe and go off of the worst-case scenerio ev%.
Weighted Average Devig Method:
- This method will always use weighted average of the implied probabilities of any of the 2 or 3 or 4 main devig methods (set in the settings). This way you can use a combination of percentages that is (hopefully) more accurate/optimal than a single devig method, and less conservative than the Worst-case Devig method.
Why are there different devig methods and which one is best?:
(Note: The below is based on my understanding and under the assumption the sportsbooks would do what I would do if I were them)
- Longshot bias. In order to best explain, let us see it from the point of view of a sportsbook:
Across all devig methods, let us always assume the sportsbook runs its analysis of the game and their computer spits out the true/fair odds are for the Steelers to have a 70% chance to win. This is true/fair odds and is 70% without any juice involved.
And let us also assume, for simplicity, that the total money wagered on both teams (combined) is $100. And let's assume the sportsbook has 8.3% juice in the market.
Steelers vs Browns moneyline. The sportsbook thinks the Steelers have a 70% chance to win (browns a 30% chance). We are the sportsbook and we want to make the same amount of money, no matter which team wins. We must adjust our odds/payouts based on how much money is on the Steelers vs how much money is on the Browns.
(Multiplicative):
In a perfect world, we would assume 70% of the money would go on the Steelers, 30% on the Browns.
This is the multiplicative method's way of thinking.
As a sportsbook, we would set the odds to:
-314 = 1.32X payout multiplier
+208 = 3.08X payout multiplier
-314/+208 = 70% fair value devig multiplicative
If we know that 70% of the money is on Steelers, 30% is on the browns, that means:
If the Steelers win, we get = ($70 + $30) - ($70) * 1.32 = $7.6
the sportsbook made $7.60 off of the game/bets. $7.6/$100 = 7.6% ROI
If the Browns win, we get = ($70 + $30) - ($30) * 3.08 = $7.6
the sportsbook made $7.60 off of the game/bets. $7.6/$100 = 7.6% ROI
(Formula = ($TotalMoneyWageredOnSteelers + $TotalMoneyWageredOnBrowns) - ($TotalMoneyWageredByWinningBettors) * (PayoutMultiplierForWinningBettors)
The sportsbook adjusted the payout/odds so that they would make the same no matter which team won. As it was a 70% chance for Steelers to win and 70% of the money was on the Steelers, the sportsbook did not need to take into account longshot bias, as there was none in this case.
What if there was longshot bias and the sportsbook used the same odds?
Let's see what would happen:
Longshot bias says that more money will be placed on heavy underdogs because it's like a lottery ticket and people love risk and a chance to win it big. Because of longshot bias, in our example below, 68.46% of the money is bet on the Steelers and 31.54% is bet on the Browns.
So, more money was bet on the longshot/Browns than in our example above.
What if we don't change the odds/payouts, even though we know that more money is on the Browns?:
If Steelers win, ($68.46 + $31.54) - ($68.46) * 1.32 = $9.63
If Browns win, ($68.46 + $31.54) - ($31.54) * 3.08 = $2.85
As a sportsbook, we want to make the same profit no matter what, and maximize our longterm profit.
So, we must adjust the odds/payouts to take into account the longshot bias (and more people betting on the underdog).
As a sportsbook, we would set the odds to:
-287 = 1.35X payout multiplier
+193 = 2.93X payout multiplier
-287/193
If we know that 68.46% of the money is on Steelers, 31.54% is on the Browns, that means:
If the Steelers win, we get = ($68.46 + $31.54) - ($68.46) * 1.35 = $7.579
If the Browns win, we get = $68.46 + $31.54) - ($31.54) * 2.93 = $7.5878
The sportsbook made the same profit and ROI% no matter which team won.
Back to our perspective as sports bettors:
We cannot see exactly how much money is wagered on either side of a bet like the sportsbooks see.
We can estimate, however, and by using the Multiplicative devig method in our first example (when there was no longshot bias), it would return the same fair value (70%). And by using the Additive devig method in our second example (when there was longshot bias), it would also return the same fair value (70%).
The right devig method depends on how much longshot bias there is in the market.
So, when is it best to use the different devig methods? It really depends. The worst-case devig method will be the most conservative in estimating longshot bias. But ideally, if we had enough data, we could estimate the longshot bias based on a number of factors such as:
market type, sportsbook, sport, game popularity, etc.
There is essentially a range of where the fair chance could be and worst-case method will give you the lowest of the estimates.
I personally prefer the worst-case method but I hope that with my new scraping site, we will eventually have enough data to analyze and determine what the true fair chance of a bet are to hit based on the factors mentioned above.
Just the Tips:
i like turtles
Patch Notes:
2/28/2022
- Added feature to allow us to specify the amount of juice in a market to calculate fair odds from that, as well as a feature to derive juice from another market.
- Changed default Kelly settings to be more normal.
- Added a cookie to save Kelly's settings
- Added some more tips
3/3/2022
- Added feature to handle mutually exclusive events (XOR)
- Added feature to handle indepdent events (OR)
3/5/2022
- Added the Combo Breaker to the site
3/7/2022
- Gave webpages titles to make them easier to switch between tabs.
- Gave feature to allow simply placing fair value into devigger, without having to place a "/".
3/11/2022
- Added the Additive devig method
3/13/2022
- Fixed a bug that caused odds to be off when using ^ and the new additive method.
3/29/2022
- Added the Worst-case devig method
- Added the Power devig method
- Improved output of combo breaker to give player name and market type on the same line, per player.
- Fixed recursion bug that would cause Combo Breaker to crash with 4-5+ inputs. It now can handle 4-5 Players/inputs. I will improve its efficiency to handle even more inputs in the future.
4/2/2022
- Added settings to customize the Worst-case method.
- Fixed a bug in the Power Method when using the "^" feature. Works now.
4/4/2022
- Added the Weighted Average method.
4/5/2022
- Fixed a bug where the Weighted Average method would only use Multiplicative if the Worst-case method's Multiplicative checkbox was checked.
4/11/2022
- Improved the accuracy the of Power method. There was a bug where 295/-400, 300/-400 and 305/-400 would return non-linear fair percentages. They were very close to eachother but because of a bug in the iteration in finding an approximate "k", it would think 300/-400 was .3% more likely to hit than 295/-400. The bug is fixed and it is now more accurate.
6/12/2022
- Added the "Show EV Color Indicator" checkbox. This lets you see red for negative EV and green for positive EV.
6/14/2022
- Fixed a bug that caused it to show Leg#1 for every leg. This bug was introduced because I am working on a custom name feature for each leg (along with much more) and didn't finish it, yet.
7/10/2022
- Added "Show All" devig method. This lets you see all 3 main devig methods at the same time.
- Added the "Show Detailed Devig Info" checkbox. This can show individual leg's juice.
7/11/2022
- Added the Additive and Power method to the Combo Breaker. I created Combo Breaker before I learned of the other devig methods. Combo Breaker had multiplicative method baked-in.
- For combo breaker, added the ability to copy/paste FD one way alt lines that are formatted like, "Player Name 3+ Strikeouts". This works for things like FD strikeout alt markets.
- For combo breaker, added the ability to specify juice in the market, for any lines that are one way. To do this you can either specify the exact juice or provide a 2 way odds for the combo breaker to derive juice from, for you. Provide a line with either of these 2 formats:
juice=7.1%
juice=-172/+134
7/12/2022
- Added the Shin devig method to the devigger
- Added the Shin devig method to the devigger
8/31/2022
- Added a section to this help file called, "Why are there different devig methods and which one is best?"
9/15/2022
- Added correlation textbox/option.
- Added boost option.
9/16/2022
- Added ability to now type parlay odds in "Final Odds" textbox and the devigger will calculate the Final Odds for you.
For example, if your parlay is for Penn State at -210 and Georgia at -400, simply type into the Final Odds box:
-210,-400
It will output -118 as the Final Odds and use those odds on the +ev calculations.
9/29/2022
- Fixed a bug in the correlation textbox's formula. It can now correctly handle 4+ legs.
10/3/2022
- Fixed a bug in Combo Breaker. It was putting too much value to the lowest alt line.
It now estimates much more accurately.