IMO I agree with you this player trend resonates better than perhaps some others. And it might be as simple as that. KG is rested and he played well so we feel he may play well next game and play a good game on D.
I enjoy your write ups and I know there was an NFL under in a Redskins game you wrote up last year that I almost couldn't finish reading all of it because I already half had my eye on the screen and was betting it I agreed so much with it. And it won.
Totally just my opinion but if you can bridge your data skills and what a fair line should be for it I think you'll do well. For instance with these skills you seem uniquely adept to be able to research particular matchups of players, lineup changes, pace of games, referees, impact of reserves on quarters or and full games, etc.
In other words, with the data work you are obviously so good at you should be able to determine as well as most what the line should be for the total. From there you can use your additional insights to refine the line to your player based trends provided those trends have meaning. I do think some of the ones you may be using aren't meaningful though.
I guess my challenge to you is that your methods are assuming that oddmakers and bettors know where the line should be. I think if you are going to involve yourself with data you need to invest the time to investigate pace of games, scoring efficiency, player rotations, etc. to come up with a number for the total. And then perhaps refine it further for some of the meaningful player trends you have found based on the previous game or the situation to make a decison on whether you want to bet or not.
I have spent a few threads knocking Dr. Bob but his models first goal is to determine a fair number. From there it is to analyze situations which may indicate which side has a greater probability of winning. The situations even if great aren't so good if the fair line is off and biased against the team in the great situation. I think his model has merit.
You have some great commentary but perhaps you're making use of situations exclusively to predict the outcome without a determination if the line is fair to begin with. And then even within that some of your "situations" I question if they truly are "situations" that are meaningful to the game at hand. That will have some trial and error. And no doubt I would question many of Dr. Bob's assumptions and situations as well FWIW.
This is a great year for NBA totals. I pointed out in a thread a Heat/Bobcat game moved 10 pts. Another Bobcat game moved 10 pts the other night from 190 to 200 and flew over. These are 20 star best bets. I believe it was you who wrote up a schedule analysis that I haven't even looked at but one figures there might be something here with totals and pace on games played by teams in the 5the game in 7 nights or 7th game in 10 nights. You won't have enough data so final decisons can't be made based only on data.
I do like the KG trend but I defintely think unless you can rationally explain some of your played based trends as to why it would help explain the total for the next game I'd be extremely hesitant using it.
