In reviewing the DLR it seems that there is money to be made in hockey totals but I am confused by the grading in DLR. There are 1112 game decisions and only 756 total decisons. I am assuming that the remaining 356 represent ties? If someone could confirm that would be great.

Those would be games where the total was either a push, or there was no AccuScore prediction. For example, if the total was 6 goals, and AccuScore predicted 6 goals, there would be no play.

Tim-
Take a look at the NHL archive. Out of the 1114 games, 358 have the same number for Vegas OU and Sim OU, and show no result as R or W. However, under OU units, these games are graded as 195 wins and 162 losses for a net +1680 (game 1303 is the only push). I think that's the question and I think the error is just in this column because if you take these games out, you're left with 416 wins and 340 losses for net +4200, which is what is shown on the summary.

Tim,

Accounting on hockey seems to be poor. The complete historical archive had the under/over profit at +5800 as opposed to the +4200 in the DLR-NHL. Also in some instances you show a 0 result for games where Simulation differed from Vegas OU (GAME ID 210 is an example). Also in the NHL Live Betting Predictions page there isn't a simulated total goal # but rather an indication based on % of whether game will remain over the total or stay below. Is grading based on higher %? Most seem to be less than 50% either way.

Thanks for helping me understand.

Pete

The records on the NHL0708 tab of the Game Archive Excel sheet contains some errors. We're currently working on fixing those.

Using GameID 210 as an example, the OU_SIM was 7, while the Vegas line was 6. However, the game was 5-1, which meant that no matter which side you took, the game would be a push, and thus there would be no chance of a win or a loss. The error in this case is the +100 that is displayed, and that is what we are working on fixing.

There are two ways where there would be no AccuScore play. One is the example above in Game 210. Another would be like Game 204, where the AccuScore simulation was 6, and the Vegas line was 6.

As for the live betting page, using Pittsburgh and Ottawa as an example, the Over is 39% and the under is 49%. That means there is a 49% chance of the game going under, a 39% chance of the game going over, and a 12% chance of the game finishing as a push.

Also, my personal opinion on the games this weekend:

The rink size in Sweeden is smaller than the NHL rinks. A smaller rink means less ice to work with, which tends to lead to more defense. The neutral zone will lose four feet in the Penguins game, compared to what the ice would be if this was played in Pittsburgh or Ottawa. Then you've got jet lag. My experience with jet lag is that it is always worse going from West to East. When I flew back from Australia to the East coast, I didn't get over the jet lag for a week. The Penguins and Senators arrived on Sunday, and will be playing six days after arriving. Assuming some players are still a bit jet lagged, that would mean that they will be moving slower, and won't be on their A-game. Finally you factor in this is the first game of the season, in an arena that is unfamiliar to both teams, and the fact that AccuScore favors the under, not taking in to consideration the above variables, and I like the under in the games this weekend.

Thanks Tim!

The DLR shows a no prediction on the O/U for Pitt/Ottawa game so for grading purposes is this a no play? I would figure that given the under prediction is 49% and over is 39% the play is on the under and would be graded that way. Sorry to be a pain but I just want to be sure I am using what seems to be a very poweful tool in the proper manner.