This article deals primarily with the Ontario (OLG) Pointspread game. Some of this could be applied to other provinces but because of the unique payout structure used by the OLG the player can significantly improve his chances of winning by understanding the payout structure and taking advantage of the opportunities it offers.
I've broken this down into four sections. It may seem like a lot to wade through but this material is very basic and an understanding of the payout structure is an absolute must if you want to be a winning player.
|PARLAY ODDS||PER GAME ODDS|
|Size||Odds||Win Prob||Win Expect||Req'd Win Pc||Odds||Win Expect||Req'd Win Pc|
|2||2||1 / 4||-50.0%||50.0%||1.41||-29.3%||70.7%|
|2 Box||2.5||1 / 4||-37.5%||40.0%||1.58||-20.9%||63.2%|
|3||5||1 / 8||-37.5%||20.0%||1.71||-14.5%||58.5%|
|4||10||1 / 16||-37.5%||10.0%||1.78||-11.1%||56.2%|
|5||20||1 / 32||-37.5%||5.0%||1.82||-9.0%||54.9%|
|6||35||1 / 64||-45.3%||2.9%||1.81||-9.6%||55.3%|
|6 Box||37.5||1 / 64||-41.4%||2.7%||1.83||-8.5%||54.7%|
|7||75||1 / 128||-41.4%||1.3%||1.85||-8.5%||54.0%|
|8||150||1 / 256||-41.4%||0.7%||1.87||-6.5%||53.5%|
|9||300||1 / 512||-41.4%||0.3%||1.88||-5.8%||53.1%|
|10||400||11 / 1024||-41.4%||1.8%||1.89||-5.6%||52.9%|
|11||600||12 / 2048||-54.6%||1.3%||1.85||-7.4%||54.0%|
|12||1000||79 / 4096||-44.8%||3.5%||1.89||-5.6%||53.0%|
The above chart breaks down the payout schedule for the OLG Pointspread game. Pay close attention to the last section, 'PER GAME ODDS'! Notice how the payout odds per game, and as a result the required win PC or break-even point, changes depending on the number of games played in the parlay. This would not be the case if you broke down a Proline parlay and is why Pointspread parlays have to be handled differently.
(with thanks to MattyKGB of the forum for correcting the 10-11-12 team values)
The above chart with some additional analysis and corresponding charts for the other provinces can be found at Pointspread Odds
So here's the first segment on how to take advantage of the payouts on the OLG Pointspread chart. This one is very basic and in time most players work this out by themselves or have the strategy pointed out to them. You'll notice that in the above chart that several of the parlay options have two entries : 2 and 2Box, and 6 and 6Box. The 'Box' refers to the strategy of playing a 2 team parlay but selecting a third game and playing two tickets, each ticket having a different side of the third game. This simple trick will signifcantly improvement your payout with no additional risk.
- Let's take a look at what happens
- Your first play is game 12 at home - 12H
- Your second play is game 13 at home - 13H
- And you're going to play the following 2-team parlay 12H - 13H for $20. Payout if wins = $40.
- Now we're going to add in both sides (box) of game 14 and split the total wagered over the two 3-team parlays.
- 12H - 13H - 14V for $10. Payout if wins = $50
- 12H - 13H - 14H for $10. Payout if wins = $50
A couple of things should be obvious in the above example. With both methods you need 12H and 13H to hit in order to cash. If both come in then one of the 3 teamers is a winner and the other is loser. By both methods your investment is the same - 20$. But if you win the second method (box) has increased the total payout from $40 to $50.
The above trick works for OLG Pointspread on both the 2 team and 6 team parlays. If you checked yourself to see what happens with other parlay sizes you would have found no improvement in payouts. However, something to bear in mind, with 3,4,7 and 8 team parlays boxing a game does not improve the payout, but it doesn't hurt either. For example a 3 team parlay pays the same with or without an additional game boxed in. In general you would of course just play this as is and save the extra work required in boxing in a game, but in the last section 'Boxing for a push' we'll come back to this with some practical applications.
Now let's move on to a more difficult area - how many games to play on a ticket.
I've made numerous posts on the online forums in which I have stated my opinion that playing 3 team pointspread tickets is a mistake. A mistake which I feel almost guarantees that the player will in the long run lose - regardless of how good a capper he may be.
There is a simple fix - play more teams per ticket. Play 4, 5 even 7-8-9 teams per ticket. It has always been an accepted belief that the best choice was to keep the number of plays to a minimum per ticket. This has sort of become an accepted rule, and the concept that more plays actually improves your chances of winning is a hard one to sell.
The math required is not that difficult and I'll go through it in this article and I welcome any questions you may have on any of the points.
For this purpose I'm going to use the posts between myself and PQ and several others from an online discussion. Nothing here is intended as a negative comment on either PQ or his plays (which in fact I agreed with). But because they are solid plays they serve to highlight very well the improvement that can be achieved by setting up the parlays as efficiently as possible.
Here are the plays and the combos as orginally played --keys Kansas St - 2.5 UCLA -0.5 --with 1 of South Carolina +1.5 Arkansas +0.5 Colorado -0.5 Villanova -1.5
This gives us 4 three team parlays. The problem is that the rotation used is in itself a huge error and changing to something like a 4 of 6 or a 5 of 6 rotation would be a big improvement.
PQ picked these plays based on the discreptancy between the offshore line and the OLG line. The method promoted on this web site. For the purposes of the thread I've added some additional information to the above.
33v Kansas State -2.5 (-4.0 -110) 57.3% 42v UCLA -0.5 (-2.5 -104) 57.0% 34v SoCarolina +1.5 (-1.5 -110) 57.0% 36h Arkansas +0.5 (-1.5 -108) 54.2% 48h Colorado -0.5 (-2.0 -110) 56.1% 32v Villanova -1.5 (-2.5 -105) 53.8%
The value in brackets is the Pinnacle line for the game. The last number is the probability of the pointspread play winning because of the weak line used by the OLG. I want to stress that this is not about how the plays were made or if you agree with the percentages given. We are going to use them for the math, but any method of selecting plays or setting those percentages would lead to the same conclusions.
its an edge no matter how many i play, when UCLA is actually a fav and im getting them straight up same with kstate they r favs by 4 to 5 in reality and i get them at 2.5 ill take it call it what u want.
I suggested that if played as 3 teamers that these were not really edge (or +ev) plays at all. I think PQ's comment on that is
more or less the accepted opinion.
After all it's either a good play or isn't, right? What difference does it make if I play it in 3 teamer or a 4 teamer?
Well that's the whole point of the matter. Strangely enough, whether or not something is a good play, the answer can change depending on how it's being played.
It obviously wins over 50% but that is not what makes a good play. In order to be a money maker
it needs to win often enough to cover the vig. The question can be answered by
calculating what win % is required to hit a break even point.
First we need to determine the OLG payout on the game. A 3-team 5.00 payout breaks down to 1.71 per game.
To double check : 1.71 * 1.71 * 1.71 = 5.0
If we recieve a 1.71 payout we can next calculate what percentage of the time we need to win in order to break even. That works out to 58.5%
To double check : if we made the play 100 times (total invested 100) we would get back 58.5 * 1.71 = 100 for even.
It should be clear now that we do not have a good play as UCLA at 57.0% does not meet the break even point. In fact playing UCLA gives an approximate loss rate of 2.5%.
Now let's try it again. Only this time we're going to play 4-teamers
Again we determine the OLG payout on the game. A 4-team 10.00 payout breaks down to 1.78 per game.
To double check : 1.78 * 1.78 * 1.78 * 1.78 = 10.0
Now using the 1.78 payout we again calculate the percentage of the time we need to win in order to break even. That works out to 56.2%
To double check : if we made the play 100 times (total invested 100) we would get back 56.2 * 1.78 = 100 for even.
Is UCLA a good play? Since their win probability of 57.0% is higher than the break even point
of 56.2 the answer is now yes.
Playing this game gives an expected profit.
So as you can see the answer to the question of 'is UCLA a good play?' actually changes depending on the number of games played. This will of course not usually be the case. But one would always receive a higher expected value by playing more games.
if i change it to 4 of 5 rotation i have to win 4 of 5 games to turn a profit......i say it is much harder to do on a regular basis than going 3 of 5
An oft cited point is the obvious one that by playing 4 teamers instead of 3 teamers you will win less often. I get it - I too would prefer as high a rate of winning tickets as possible.
But the important issue is not how often you will win, but will you win often enough so that the payouts will cover the losses and leave a profit.
And to that question it is quite possible that using the same plays the answer will be no with 3 teamers and yes with 4 teamers.
Again the plays listed in the original post are perfect to demonstrate this very clearly. I've used the plays above with their win probabilities and entered them into my ParlayMaker program. I could go through the math here but it would take a bit of time and my program will spit out the final answers quickly.
First here are the parlays using the key the first 2 and add 1 rotation : 33V - 42V - 32V 17.57 -12% 33V - 42V - 34V 18.62 -7% 33V - 42V - 36H 17.70 -11% 33V - 42V - 48H 18.32 -8%
Now here's the parlays using a simple 4 of 6 rotation 33V - 34V - 36H - 32V 9.52 -5% 33V - 34V - 36H - 48H 9.93 -1% 33V - 34V - 48H - 32V 9.86 -1% 33V - 36H - 48H - 32V 9.37 -6% 33V - 42V - 34V - 32V 10.02 0% 33V - 42V - 34V - 36H 10.09 +1% 33V - 42V - 34V - 48H 10.44 +4% 33V - 42V - 36H - 32V 9.52 -5% 33V - 42V - 36H - 48H 9.93 -1% 33V - 42V - 48H - 32V 9.86 -1% 34V - 36H - 48H - 32V 9.32 -7% 42V - 34V - 36H - 32V 9.47 -5% 42V - 34V - 36H - 48H 9.88 -1% 42V - 34V - 48H - 32V 9.81 -2% 42V - 36H - 48H - 32V 9.32 -7%
And lastly the same plays using a 5 of 6 rotation 33V - 34V - 36H - 48H - 32V 5.34 +7% 33V - 42V - 34V - 36H - 32V 5.43 +9% 33V - 42V - 34V - 36H - 48H 5.66 +13% 33V - 42V - 34V - 48H - 32V 5.62 +12% 33V - 42V - 36H - 48H - 32V 5.34 +7% 42V - 34V - 36H - 48H - 32V 5.31 +6%
You can see the difference.
The 3 teamers are going to result in an expected loss rate of about 10%. None of the parlays are +ev. The 4 teamers do much better. Most of the expected results are hovering close to the break even point. And with the 5 teamers the gain is huge. You've now changed from a player who has an expected loss rate of 10% to a player who will win at a 10% rate. Make plarlays like this all year long and you will end up with some of OLG's money in your pocket. I think you'll agree that using the same plays and only changing the rotation used, the improvement is quite dramatic.
Again I want to stress that the above only applies to OLG Pointspread. Not proline, not offshore, and not even pointspread in other provinces.
The reason why this works is that the OLG's payout chart is screwed up. How they came up with it I have no idea but you can take advantage of its flaws by simply increasing the size of your parlays.
If you worked your way through the last section then congrats, that was the most important part and we're almost at the end.
You'll recall from earlier that when playing a 2 or 6 team parlay the correct strategy was to box in an additional game to improve your payout with no additional risk. Now we're going to extend that strategy taking advantage of the ability to box 4,5,7,8 team parlays with no penalty on the payout.
I'm going to demonstrate our revised playing strategy with an example.
- You have 3 NFL games you're going to play
- 11H Pittsburgh -1.5
- 12H New Orleans +4.5
- 13H New England -7.0
We could play $20 for a payout of $100. Or we could box in a 4th game and play two 4-team parlays for $10 each. Each of these 4-teamers would have a payout of $100. If our original 3 teams come in we would once again collect $100.
Same thing, so why bother? Almost the same, but not always. Take a look at one possible outcome : 11H winner, 12H winner and 13H New England wins by exactly 7 - push! Now what happens to our parlays? Our 3-teamer now gets reduced to a winning 2-teamer with a reduced payout of $40. But our winning 3-teamer+box now gets reduced to a winning 3-teamer with a reduced payout of $50!
Now if you understood the previous section on how many games to play, you're probably not going to be playing any 3 team parlays. But this trick also works with 7 team parlays, which could reduce to 6-teamers or 6-teamers+box. You may think that this is an unlikely occurrence. But trust me on this, if you play a lot this little trick will make you extra money over the course of the year.
In the above example I used an NFL game, but it should be obvious that if playing MLB Pointspread, that this becomes even more important. In games were one side is favoured by 1.0 the chance of a push is large!
My recommendations on this. In football and basketball if playing a 3 or 7 team parlay, if any of the games in the parlay could result in a push, I always box in an additional game. In theory other variations on this theme are also possible but whether or not it is worth the time involved for a very low probability event is up to you. For example, if two games could result in a push then we could box 2 games in a 7 teamer to create 9 teamers. Or boxing a game in 4 or 8 teamers could also be optimal.
In MLB pointspread I box games even more. For example, I will add in a boxed game to all 7-teamers, even if there are no push lines, in case of a rainout. If there are one or more potential push lines, I will also add a boxed game to all 4 and 8 team parlays. Between postponed games and pushes, reduced parlays are very common in baseball - be ready for them!
If you made it all the way to here, then congrats, you are now an expert on the twists involved in setting up pointspread parlays. Any questions on any of the above can be directed to me on the forum.