| 0 |
Subject: Football Pickoff
Posted by: Guru
- [330592710] Tue, Aug 27, 2002, 16:33
I've launched the 2002 version of Football Pickoff. I'll make the formal announcement tomorrow. Meanwhile, I'd like to have some of you kick the tires today, just to make sure everything is working properly, and there aren't any obvious bugs, bad links, typos, grammatical faux pas, etc. Please report any issues here.
Assuming there aren't any big problems, this is live, so when you register, it is for real - not a test.
Thanks. |
| 1 | leggestand
ID: 501029817
Tue, Aug 27, 2002, 17:48
|
Guru,
I dont know if this is a glitch, but I made my picks, and some registered in the %'s for games, while others did not. I.e., my SF vs NYG registered and my Pitt vs NE did not. My user name is leggestand.
|
| 2 | Guru
ID: 330592710
Tue, Aug 27, 2002, 17:53
|
Perhaps the routine that updates the percentages doesn't handle a 16th game. I'll take a look. It appears that the choice is saved, but the tallies just aren't active.
|
| 3 | Guru
ID: 330592710
Tue, Aug 27, 2002, 17:56
|
I think it's OK now.
|
| 4 | Sludge
Sustainer
ID: 566332517
Tue, Aug 27, 2002, 18:42
|
Well kick me in the head! I just spent a few hours today tweaking my Pickoff program and updating it to handle 16 games. I had to re-learn how the algorithm works by reading my completely comment-free C code. Ugh, what a headache!
Not a big deal, but the rules still mention that the maximum bonus points available is 300. Of course, that's 350 now with 16 games.
Baseball players can go and stay on strike as long as I've got Pickoff. :)
Thanks again, Guru. Hopefully I can reclaim the top spot.
|
| 5 | Sludge
Sustainer
ID: 566332517
Tue, Aug 27, 2002, 18:49
|
Guru -
Possible bug. There is no default to "Skip" on the last game.
|
| 6 | Ender
ID: 13443221
Tue, Aug 27, 2002, 21:09
|
Hey, Sludge, how'd you finish last year?
|
| 7 | Stuck in the Sixties
Leader
ID: 207362721
Tue, Aug 27, 2002, 21:40
|
Don't seem to be able to login, even after re-registering. Could it be I'm using my computer at work?
|
| 8 | beastiemiked
ID: 587492117
Tue, Aug 27, 2002, 21:42
|
Sludge, I'm pretty proficient in C. Why don't you just send it to me and I'll fix it ;-)
|
| 9 | Sludge
Sustainer
ID: 566332517
Tue, Aug 27, 2002, 21:42
|
Three words:
Not too good!
So bad, in fact, that I don't even remember approximately where I finished.
|
| 10 | Sludge
Sustainer
ID: 566332517
Tue, Aug 27, 2002, 21:47
|
beastiemiked -
Already fixed. Thanks for the offer, though. Sarcasm it was, but did you know that I already made it publicly available here? Of course, it's just an executable, but you can reverse engineer it if you want. Just don't try to run it with 16 teams.
|
| 11 | Sludge
Sustainer
ID: 566332517
Tue, Aug 27, 2002, 21:50
|
Ahhh... there we go Ender. Top finish was 157th. Other teams were 179th and 333rd. The 179th place team was actually trying for last place, and it clobbered my 333rd place team.
|
| 12 | beastiemiked
ID: 587492117
Tue, Aug 27, 2002, 21:51
|
Cool. I just always thought you kept it to yourself as you are always at or near the top at the game.
|
| 13 | Guru
ID: 330592710
Tue, Aug 27, 2002, 23:00
|
Sludge[4] - rules updated to refer to max bonus of 350.
Sludge[5] - bug fixed.
Stuck[7] - your registration looks OK. I was able to log into your account. Successful logging requires cookies and scripting enabled. Perhaps your work computer has some sort of security feature that's interfering?
|
| 14 | Ender
ID: 13443221
Tue, Aug 27, 2002, 23:07
|
Thanks, Sludge. I was just curious because I know how much work you put into it.
I intemd to stick it out and play the whole season this year. I gave up early last year when my 8 year old was doing better than me :(
|
| 15 | Guru
ID: 330592710
Wed, Aug 28, 2002, 08:43
|
Sludge (and others):
There is one feature of the Pickoff scoring method that has always bothered me. If you get a negative score in week #1, you are better off abandoning your entry and starting over, because you're already behind everyone who didn't even play the first week. For that matter, people who participate in any week and get a negative score do worse than those who sit out.
I don't think there have been many, if any, who have taken advantage of this implicit mulligan option, and those who have competed well have generally played evey week. So I don't think my concerns have had much if any impact in the ultimate standings.
But I've been thinking about two scoring alternatives. I won't be implementing either this year, but I might be tracking them for future consideration.
Option #1: If, at the end of any weekend, your cumulative score is negative, your score is automatically reset to zero. This is the same as saying that once you've gone negative, you get to throw away your entry and start fresh.
Option #2: Game scoring is the same as usual, but the total score for a weekend cannot be negative. Thus, if your game scores tally to a negative number, your score for the weekend is simply zero. Your cumulative score is the sum of your positive weekend scores. This is a more dramatic change, but also potentially simpler to explain and understand.
I assume that either one of these options could theoretically change the optimal strategy, although I'm not sure just how. In the first case, the strategy could change at the outset, but once you had built a positive result, the game reverts to the usual game. In the second case, the optimal strategy would generally be altered throughout.
The reason for the strategy shift is that the risk/reward tradeoff is altered, since an excess of negatives essentially are forgiven. Thus, the potential downside of a bad weekend is not nearly as severe. What I can't easily assess is just how this might impact the selection strategy. (My guess is that it would not change the approach that most players take. But it might very well impact the picks that an optimization routine would make.)
Any thoughts? (Sludge - this type of anaylsis seems right up your alley!)
|
| 16 | Sludge
Sustainer
ID: 113368
Wed, Aug 28, 2002, 09:21
|
Guru -
As to the mulligan "option", I never took it because I always thought we were limited to a maximum of three registrations in order to qualify for the prizes. I believe the second year, I held an entry back a week to see how my others went that first week, and I expect I'll be doing that again this year. Unless, of course, it is acceptable to abandon an entry with negative cumulative score and start a new one.
Option 1 doesn't alter the optimal week-to-week strategy. That remains the same. It may alter the way you approach the game in the beginning since even the "optimal" strategy has a good (1 in 4?) chance of going negative. You have to weigh the chance of the active team going positive and how much the active entry can score against the potential score of the non-active entry. In other words, I may actually be hurting myself because the non-active entry could very well have gotten me a nice positive score and it's sitting at 0 instead.
An idea, though, is to take what you think is your "best" entry and just pick the exact opposite for another entry. That way you're guaranteed to start off with at least one positive entry. Of course, it also guarantees that one of them will be negative at the start as well.
As to Option 2, it doesn't change the optimal strategy. The optimal strategy is to maximize the expected points based on the probability you believe each team has of winning. All you're doing is replacing those entries with negative expected points with zeros. Since the maximum is always positive, it still remains optimal. In addition, you would tighten up the races for the top since you lower the variation in week-to-week scores.
If you want an option which probably will change the optimal strategy, you can add a number to the score which will guarantee a minimum possible score of 0. Once you've selected the teams you want for your entry (say, N of them), you lock yourself into one of N+1 possible scores. All you would have to do, Guru, is compute the lowest possible score for each entry (it's easy to do), and add to that entry a number of points equal in magnitude to the lowest possible score (which is negative). My guess is that this would probably alter the game to a point that nobody would like. As a first guess, I would say that the optimal strategy would be to pick all favorites. I'll change up the program (won't be hard) and run it for some test cases and see what would happen. My reason for guessing is that you would take away the risk of picking favorites, and actually start them out on a higher plateau since their lowest possible score is much larger in magnitude than if you were to pick all dogs.
|
| 17 | JeffG
Sustainer
ID: 40451227
Wed, Aug 28, 2002, 10:36
|
RE [15] The few that do mulligans if they finish negative week 1 are fine by me and probably just to do nothing is ok. Alot of debate for so few.
My vote is for as is.
Otherwise, I like option #1 and just have the system do auto mulligan for everyone ending a week in the red so no one takes advantage of this little loop hole.
My issue with option 2 would be that it will make someone's lead a little more insurmountable in the post season with fewer games where it is more likely to yield a few nagatives (especially the Super Bowl). If you went option 2, wave it for the post season.
However a few brainstorming ideas to shoot down, neither of which I like better then what I said above:
- Giving everyone a 500 point base for picking each week and have the scores for the week go up or down from that point.
- Allow everyone to drop their worst (non post season) weekly score if they enter week 1. Those doing the mulligan or those joining week 2 (or later) would have a zero for week 1 which would be the one dropped only if they went the whole season without a negative week.
|
| 18 | Guru
ID: 330592710
Wed, Aug 28, 2002, 10:49
|
Sludge - you are right in the 3-entry limit effectively precluding massive mulligans. But even so, I don't like the concept that you can be better off by skipping the first week than by playing poorly.
Second, I'm really not looking to shift the optimal strategy. (I'd really prefer not to... "if it ain't broke...") I was just wondering whether tracking the results using either of these options might be misleading, since game play would probably be different if altered scoring rules were in place. My suspicion was that, in practice, any shift in strategy would be minor for these potential changes. Maybe even negligible. And your analysis suggests that any strategy shift is possibly irrational, or at least sub-optimal.
I think I'll track the results under Option #2 this season (negative weeks are set to zero), and see how that changes the results. We can decide later whether that's something worthy of consideration for next year.
|
| 19 | StLCards
Sustainer
ID: 4431816
Wed, Aug 28, 2002, 11:07
|
If you still had all the picks from last year you could pick a random sample and see how the different scoring options would have worked.
|
| 20 | Seattle Zen
Donor
ID: 554192913
Wed, Aug 28, 2002, 11:26
|
Option #1 is a good idea, option #2 is a horrible idea that completely changes the game.
The way to score the most points in this game is to pick every game correctly and garner a 350 point bonus. The most logical way of doing that is to pick favorites. With no negative -500 point week looming for that strategy, favorites would become even more popular, ruining the unique balance that I enjoy in the game's present incarnation.
My strategy is to conservatively pick each week to limit my exposure to a huge negative week. It is fun watching the teams ahead of you pull a -400 week while you stay steady with a contrarian +50 or better. When you pick a lot of dogs, you are not saddled with as large a negative as you would be when you pick a lot of favorites and the dice end up snake-eyes.
|
| 21 | leggestand
ID: 501029817
Wed, Aug 28, 2002, 11:30
|
I agree with Zen. There has to be a penalty for having a bad week. People can pick the favorites each week, and not be penalized by a negative score. I would go with option 1.
|
| 22 | tduncan
ID: 423452813
Wed, Aug 28, 2002, 11:38
|
The negative scores are very much a part of this game. I remember last year there was a week (I think 16) were allmost all the teams around the top finished deep in the red. I finished with like -20, and ended up going some spots. If you make weeks like this all 0, then it's like this week had nothing to do with the pickoff game - everyone got 0 points. Not a good idea IMHO.
Make you could make something like this:
When somebody create a team in midseason, he automaticaly get the score of the worst team in the game, or the average of all negative scores, or the average of all teams under the median (the bad half of teams). I think something like that would be more fair.
|
| 23 | Stuck in the Sixties
Leader
ID: 12451279
Wed, Aug 28, 2002, 12:10
|
Guru:
Just tried to login from home and got the same "login failed" message.
I'm now aware of any substantial changes to my computer since last season except that I now have a cable modem.
|
| 24 | Guru
ID: 330592710
Wed, Aug 28, 2002, 12:20
|
Seattle Zen[20] - your argument is compelling. Obviously, I hadn't thought this through very much.
Maybe I should simply consider an "auto-mulligan" (still for 2003, not 2002), which automatically resets the cumulative score to zero anytime it goes negative. Beyond that, there is probably no need for a change (if there is even that need).
But maybe even that is a bad idea, as there are teams that stay negative all year. Perhaps that option should only exist for the first week, or the first several, just to get people acclimated.
|
| 25 | Guru
ID: 330592710
Wed, Aug 28, 2002, 12:22
|
Stuck - send me an email and tell me what you are entering for your login and password. I'll compare that against what is on the registration file. Sounds like there may be a disconnect.
|
| 26 | Sludge
Sustainer
ID: 566332517
Wed, Aug 28, 2002, 13:37
|
JeffG -
Option 2 would actually make it harder for someone to gain that insurmountable lead to begin with.
As to my feelings on the options, I would agree (gasp!) with SZ here on his feelings about option 2. While it doesn't change the "optimal strategy", it would change the strategy of many of the players who would start picking less dogs because there would be less risk involved.
As to option 1, I could care less. It'll help everyone stay competitive for the first couple of weeks, but after the leaders start to build large positive scores, that will disappear. It may help keep interest in the game high at the start of the season, though, and that would be a positive.
|
| 27 | Stuck in the Sixties
Leader
ID: 12451279
Thu, Aug 29, 2002, 11:14
|
Guru:
EMail sent. Just tried again to no avail
|
| 28 | Guru
ID: 330592710
Thu, Aug 29, 2002, 11:41
|
Stuck - I responded to your email earlier this morning (to both addresses). You should have my reply by now.
None of the three login names your mentioned ("Stuck60", "Stuck600", "Stuck6000") are in the registration database. But "Stuck 60" (with a space stuck between Stuck and 60) is registered.
What I'm not clear about is whether you have tried to register all three of your entries, or just the first. If it's just the first, then the space is the problem. But if you are having trouble registering the second and third, I dont understand where that breakdown is occuring.
|
| 29 | Stuck in the Sixties
Leader
ID: 12451279
Thu, Aug 29, 2002, 12:33
|
Guru:
It was the spaces. Damn computers keep on doing what you tell them to rather than what you want them to do.
Thanks
|
| 31 | root88
ID: 359101014
Tue, Sep 03, 2002, 14:13
|
Has anyone started a division that I can enter? I would like to be ranked in a smaller group as well as the whole. I would have liked to join the Delaware division since that is where I am from, but any will do.
|
| 32 | Perm Dude
Leader
ID: 87192619
Fri, Sep 06, 2002, 17:48
|
BUTT
|
| 33 | Chestergreat
Sustainer
ID: 58224208
Sun, Sep 08, 2002, 19:08
|
Great start . . .made Cleveland my Double point pick and they blow the game with two taunting penalties in the last 15 seconds.
|
| 34 | Sludge
Sustainer
ID: 113368
Sun, Sep 08, 2002, 22:43
|
All three positive scores so far. Off to a better start (knock on wood) than normal.
|
| 35 | beastiemiked
ID: 587492117
Mon, Sep 09, 2002, 00:25
|
My teams did horrible, 2 with under 50 but positive. The other -71 eventhough I got the Texan and Carolina games right.
|
| 36 | Motley Crue
Donor
ID: 42842311
Mon, Sep 09, 2002, 08:32
|
What about the people who play to try to get the lowest possible score? They would be real unhappy if negative scores were reset to zeroes. I personally would have been real happy if the Rams and Ravens would have won. I probably would not have a -85 right now if they had.
|
| 38 | StLCards
Sustainer
ID: 4431816
Mon, Sep 09, 2002, 10:07
|
Was there a worst score prize last year?
|
| 39 | Sludge
Sustainer
ID: 566332517
Mon, Sep 09, 2002, 10:57
|
No, there wasn't. Guru didn't award one because then the percentages wouldn't reflect how many people thought they were going to get points based on who won the game.
|
| 40 | sarge33rd
ID: 324532412
Mon, Sep 09, 2002, 13:06
|
-34 for me motley. at least, i know now the identity of someone i'm ahead of! lol
|
| 41 | MyNiners
ID: 13782314
Mon, Sep 09, 2002, 14:51
|
Sarge you are WAY ahead of me..I think I must have a misconception here somewhere. My score is -131. Feel my pain, I was NOT trying for a low score, I just suck
MyN
|
| 42 | Chestergreat
Sustainer
ID: 30742914
Mon, Sep 09, 2002, 16:52
|
-71 for me.
No taunting from Cleveland would have had me at +98 instead of -102 for that game . . and a nice +129 start. Oh well . . it's a long season.
|
| 43 | root88
ID: 359101014
Mon, Sep 09, 2002, 18:46
|
First time playing a pickem game. 438 points makes me very happy. I doubt it will last though.
|
| 44 | Guru
ID: 330592710
Fri, Sep 13, 2002, 12:44
|
Just for kicks, I added a real time tally of the doubled picks for each team. When you pull up the pick page, you can now see not only how many times each team has been selected, but how many times it is doubled. Ditto for the scoring recap (for prior weeks).
It's kind of interesting to see how spread out the doubles are.
|
| 45 | balls
Donor
ID: 193311316
Fri, Sep 13, 2002, 15:11
|
I don't see where that has been added, Guru.
|
| 46 | StLCards
Sustainer
ID: 4431816
Fri, Sep 13, 2002, 15:24
|
works for me
sample: Atlanta (0-1-0) 39% [99-9]
99 people have picked Atl so far and 9 have doubled on Atl.
|
| 47 | Guru
ID: 330592710
Fri, Sep 13, 2002, 15:28
|
Click on the link to "Review Score". After each team, there is a percentage, followed by two numbers in [brackets]. The first number is the number of people who picked that team, and the second number is the number of doubled picks. For example, in the first week, 259 people picked San Francisco, of which 10 had that game doubled.
The same format is also used on the "Make Picks" page. The percentages and number of teams picked have been there for awhile. The number of doubles is the only new information.
|
| 48 | balls
Donor
ID: 193311316
Fri, Sep 13, 2002, 15:42
|
there it is, thanks guys.
I like the new addition, Guru.
|
| 49 | slimer
ID: 31841143
Sat, Sep 14, 2002, 03:54
|
who cares what you think, balls ;)
|
| 50 | balls
Donor
ID: 193311316
Sat, Sep 14, 2002, 11:55
|
hey, where ya been Skippy? you ever gonna respond to me email?
|
| 51 | slimer
ID: 6847170
Tue, Sep 17, 2002, 00:58
|
what email?
|
| 52 | StLCards
Sustainer
ID: 23730222
Tue, Sep 17, 2002, 01:05
|
I think there should be an extra penalty for choosing Washington to win this week! I was hoping they would call the game when the cops sprayed the Eagles with pepper spray in the 4th quarter. That game was hard to watch!
|
| 53 | slimer
ID: 6847170
Tue, Sep 17, 2002, 01:11
|
sorry, balls...i just found it. Just emailed you back.
|
| Rate this thread: | If you wish, you may rate this thread on scale of 1-5. Ratings should indicate how valuable or interesting you believe this thread would be to other users of this forum. A '5' means that this thread is a 'must read'. A '1' means that this is a complete waste of time. If you have previously rated this thread, rating it again will delete your previous rating. If you do not want to rate this thread, but want to see how others have rated it, then click the button without entering a rating, or else click here. |
|
|
Post a reply to this message: (But first, how about checking out this sponsor?)
|