Well the other guys sure seem to be doing something wrong compared to you guys! You can flip a coin and be 7-7!! I wish we could get one final clarification on this thing.
-
Comment -
why these numbers? Arent we supposed to use the percent?Originally posted by pubster77
Just asking ....just trying to understand is all.Comment -
I knew I was right!Originally posted by porkchop817Comment -
Java Program
Has anyone been using the Java program? I entered a few in on my own and it seems to be correct.Comment -
Becareful though, it might be giving us plays that we don't want. I compared the two records, from when we started until today, and this is what i found (not including the two games from 12/15 for either team):Originally posted by Rio DiNero
In that time we made 12 plays, they made 10.
In that time our record was 10-2, theirs was 7-3.
For some reason they didn't count Phoenix last night in their record, even though alot of guys posted it as a play. The guy who is keeping the record says Phoenix wasn't a play because the line was -10.5
. I didn't see it that high anywhere. http://www.sbrodds.com/ lists the opening line as PHO -8.5, closing line as PHO -9.5.
If you count Phoenix as a play for them, the records look like this:
SBR: 10-2 .833; RX: 7-4 .636
If you look at their record for the games before we started tracking the system the other forum is 2-3 .400
This is their record for the season up until today, not counting the two 12/15 games, and counting the Phoenix game: 9-7 .563
Tread lightly until we can get a bigger sample size.Comment -
Probably just a coincidence...GM covered.Originally posted by underdogComment -
I ran the sample of George Mason at Dayton...
George Mason 8-2 SU = .800
Dayton 11-1 SU = .917
917 - 800 = 117
117/20 = 5.83
5.85 + 3 = 8.83
8.83 -9.5 = -0.67
NO Play
Remember this system began play only after 20 games had been played by each team.Comment -
Philly 76ers come out to below 0 for their upcoming game...haven't checked on injuries but this may be a playComment -
I dont know why we use these numbers, just that we do. I guess you could use the percentages and then divide by 2 instead of 20.Originally posted by underdogComment -
So far there are no plays on the teams that have lines out, there are two games that don't have the lines up, and here is what needs to happen for them to play:
NJ Nets @ Detroit
Richard Hamilton is doubtful for Detroit
Detroit needs to be the underdog by +1 or more, and they need Hamilton in the game, for this to be a play
Denver @ Toronto
Denver needs Carmelo Anthony, and Nene Hilario to start tomorrow
Denver needs to be the underdog by +2 or more, and they need the two listed players in the game, for this to be a playComment -
if it is below 0, the play would be LA Clippers +3...but the clippers would be back to back since they played earlier tonight.Originally posted by SportsLockPicksComment -
My interpretation of using the number 20 is the conditional rule that the author originally intended which was to wait until each team had played at minimum of at least 20 games before starting to use this system. This would give you a reasonable sample size of past performance to weigh the W\L data.
I am also going to use & track this formula from each team's records for:- overall last 20 games played
- overall Home and Away W\L records
- overall ATS records
- last 20 games ATS
- overall Home and Away ATS records
Comment -
no way denver is underdogged against toronto, and richard hamilton was out last game so that would still be a no play if he came back cause you need a week for starters to returnComment -
good banters fellas. i like it. keep track of whatever u see fit. share here if u see any promising results.Comment -
Will do mundane. Just have seen so many variables of this system i my web travels. Will take some time over the next week to set up different spreadsheets for all I have mentioned and will also try and backtrack thru 2007-08 and 2006-07 to see if this formula and/or these parameters were or may have been a valid and profitable approach.Originally posted by mundaneComment -
I agree with you that there is no way denver is an underdog. However, maybe I misunderstood the 1 week rule. I thought it was: if an injured player is going to be out for an extended period of time, give the team one week without him before betting on them.Originally posted by nick2060
I'll go back and review it.
edit: you are right nick
Comment -
Guys! Do I have to keep on bashing you over the head! The system is 10-2 and you only have to do some fairly simple math and apply some filters. It had a bad beat the other day on Phoenix when Nash went out early! Why do SOME of you keep trying to change it!!!???
Ten out of 12! The one thing I would like to know is how the back to back affects the system because having followed this system it seems like the back to back takes you away from some pretty good plays.Comment -
the other forum confirms that there are no plays today...including no plays with their positive/negative computation stuff.
sit back and enjoy some bowl games boys
Comment -
I agree DD as some plays were quite valid IMO but you can always alter the filters for your own preference and quantify results from that standpoint. The is a system as you stated and every time the filters are changed or adjusted, it will attain different results every time. I believe we need to look at each team's situational play results in B-B situations both as a winner and loser and then possibly apply a different filter based on those results. This entire system is still totally based upon past results and cannot predict correctly, any future event even more than say 70% of the time.Originally posted by DevilDog
It is a great tool for anyone's handicapping arsenal just like power ratings, strength of schedule, travel, extended road trips, injuries, coaching strengths, etc..
That is why I am also going to use it as a basis for not just overall Win% but look more closely at Home and Away Win% records and ATS records as well. We have a basis to work from here and time will only tell how this pans out against anyone's bottom line $$$.
Regarding the Phoenix game the other day, there was no way the system would know ahead of time that a positional team leader like Nash would leave the game and not return, thus altering the potential system play result. I treat that one as an invariable beat, not a bad beat. Many times the linemakers will adjust a game line as a result of a missing starter before the game begins. The filter regarding starters is a good one IMO and is very valid.Comment -
I think this has definite merit for NBA and hope it is something that all readers can take and build on.
I will tell you that I have spent a few of hours reviewing about 100 College games and tried out periods of the season over last 3 years after 20 games were played. The results were less than 50% no matter which way I turned and for each season. I tried the Java program and also doing it manually. Obviously whoever developed this did not recommend it for college for a reason.
I am curious though..If this is such a great tool for NBA, why is it just making it into the forum with this much discussion. That seems odd to me.Comment -
its relatively unproven.
for example I think JohnnyBaseballs is something like 13-5 in his 100 game duel with Sam n company. But would you just start following all of Johnnys plays? Its the same way for this system. 10-2 is great, but people generally want a higher sample size.
I think this system will work better in NBA because a teams record is so important in the formula. In the NBA you know the quality of opposition whereas in College teams records can be inflated due to playing poor teams. The difference between a good team and a bad team is a lot less in the NBA. ie. In college you will see games with 20 point spreads regularly.Comment -
yep!
Certainly much to learn. Hope it turns into something good.Comment -
After doing some reading at the origin of this system I found the reasoning for "divide by 20"
I think this is a good explanation. He uses 20 to divide because that is what works.
For us math nerds, it's a factor that makes the difference in winning percentage that equates it to point spreads. The higher the difference the more points a team should win by. It's a factor of giving 50 times the difference of percentage. The percentages are multiplied by 1000 before doing the subtraction and the difference is divided by 20: 1000 / 20 = 50.Comment -
Originally posted by Meestermike
You might be right meestermike but I also have a theory. There are 26 week in a regular NBA season but since we are asked to do our calculations after 20 games, that's about 6 weeks leaving you with a total of 20 weeks. But what I have been doing is dividing by 26 instead of 20 giving me a more accurate percentage to play or no play. Maybe mudamee can do a sample spread sheet with the divident of 26 instead of 20 and see what happens...Comment -
There does NOT have to be a specific reason for using 20 other than the fact that number is what brings about the highest win percentage when they tested it. These types of systems are routinely tested with pretty random numbers....not all numbers in a system have to be directly related to the point spread, win percentage, or numbers like that.
If you understand systems, there is almost always ONE number that seems to make no sense....it is the number that is found to "work" best with the other numbers in the system.Comment -
post early for the new year. what play is approve base on the system.
hous is a bum team.Comment -
Denver -7.5 at Thunder is a play!
Comment -
I wonder if there is also a system to figure out the over/under?
Comment -
To noobies: be careful Vegas is watching you
Man people are all over this thread. I wonder if when there is a "play" the bookies put less juice on the ML? hehe they've gotta be looking at this thread. It is one of SBR's biggest NBA threads now.
Be careful; here is a story passed down by Ben Graham:
"Let me tell you the story of the oil prospector who met St. Peter at the Pearly Gates. When told his occupation, St. Peter said, "Oh, I'm really sorry. You seem to meet all the tests to get into heaven. But we've got a terrible problem. See that pen over there? That's where we keep the oil prospectors waiting to get into heaven. And it's filled—we haven't got room for even one more." The oil prospector thought for a minute and said, "Would you mind if I just said four words to those folks?" "I can't see any harm in that," said St. Pete. So the old-timer cupped his hands and yelled out, "Oil discovered in hell!" Immediately, the oil prospectors wrenched the lock off the door of the pen and out they flew, flapping their wings as hard as they could for the lower regions. "You know, that's a pretty good trick," St. Pete said. "Move in. The place is yours. You've got plenty of room." The old fellow scratched his head and said, "No. If you don't mind, I think I'll go along with the rest of 'em. There may be some truth to that rumor after all."
Comment -
they use 20 because it finds the median between the 2 percentages.
like someone already said, you can just take the difference in win % (2 digit number) and divide it by 2 rather than take the difference in the 3 digit numbers and divide by 20. the 3 digit number will be slightly more accurate.
so really youre looking at the difference in win% and dividing that by the number of teams (2). the result is the difference between team a or b win% and the median of the two. that difference somehow correlates with the margin of victory between the teams. if the spread falls short of this projection, then the fav is the play and if the spread extends beyond this projection, then the dog is the play.Comment -
willpies,
play is on denver???? hope you still see my message.Comment -
yes it is. good luckComment -
thanks. good that you are on line. will give my attention on denvers lineComment -
I'll try to contribute to this thread a bit more if I can ..
NUGGETS ARE A PASS
Yes, the total after the math is +13.25, but Melo has not been back for one week (he did not play against the Sixers last Friday .. but he did play on Sunday) and Nene missed Monday's game against the Hawks (has not returned for one week).
These games haven't come out with lines yet:
CELTICS COULD BE A PLAY
Boston is a system play if they are favored by a max of 15 points. Keep an eye on this one because if Caron Butler is ruled out, the spread could be more than the key number 15. If Butler plays, the line should be within 15 if you jump in early.
ROCKETS ARE A PASS
Shane Battier is out. I would still play this if Adelman played Artest as a starter - instead, he moved Aaron Brooks into the starting lineup in the last game and shifted T-Mac over.
PISTONS ARE A PASS
Rip Hamilton is questionable, but he has missed a game this week.
SUNS ARE A PASS
Steve Nash is questionable with back spasms, but sat out the team's last game earlier this week.
BLAZERS ARE A PASS
Brandon Roy sat out a game earlier this week.
So keep an eye on Boston .. that's the only play I can see right now .. also, if the Lakers' line dramatically falls to -5, take that .. but we all know that isn't going to happen!Comment -
Thanks Indeed.Good work.
I was actually thinkin Celtics on the spread intuitively. They just had a humiliating defeat and have been away for a while. They will be home.Usually when a team comes home they crush the other guys.Comment
