The Big Card

[Vinny Vidivicci]

No idea, it's your graph.

A bug, maybe?



J/K

What's the method you and Therber have been using? Taking high % per inning teams?

~ VV

[sycoogtit]

No idea, it's your graph.

Well that's hardly helpful. I've spent a lot of time on this, and I may have saved you some money by showing that, unfortunately, the system is quite erratic. I think the least you can do is spend a few minutes analyzing the results and see if you can explain this oddity. Perhaps it will give you or someone here an idea on how to tweak this system for better results.

[Vinny Vidivicci]

I was just kidding - I can't explain it anymore than you can. We can analyze the graphs forever, but I keep coming back to the main premise of the system - which was to find "LIVE" dogs.

If you recall, in the post that I explained my theory, and why I chose an 8 game window, which was to try and FIND and BET dogs who had WON at least 5 of their last 8 (wherever they are today), and to FIND and FADE faves who had lost at least 5 of their last 8 (wherever they are today). In crunching the numbers year-to-date, which at that time was through May 15th; I began by tallying ALL teams with an 8-game-at-site "advantage" to give me a baseline of how this system was doing. It then "morphed" into what it is now, which is to take all teams (other than ADs) who had at least a 1 game "advantage".

The only thing I can do is go back to the drawing board. Here's what I'm going to do -

Not only am I going to back-check the "Take LIVE Dog / Fade DEAD Fave" idea above, but I am also going to compile ALL of the combinations of situations that occur. The only problem with this is that I am going to be drawing conclusions from a small sample size. I'll post what I come up with here when I'm done. Do you understand what I mean by all of this?

~ VV

[therber2]

big card 5/19

Indians -130 (3 units)
Rays +141 (.5 units)
Nationals -135 (1 unit)
WAS/NYM u8.5 +125 (2 units)
Phillies -130 (10 units) (wind gentle backwards)
PHI RL +145 (5 units)
CUB/PHI o9.5 EV (1 unit)
Brewers -145 (1 unit)
MIL/PIT o9 -110 (2 units)
Reds +125 (1 unit)
CIN/ATL over8.5 -105 (4 units)
Twins +110 (5 units) (wind strong blowing back)
Rockies -115 (.5 units) (gentle wind out)
Texas +125 RL (6 units)
Orioles +140 (2 units)
White Sox -150 (10 units)
under8.5 -105 (1 unit)
Cardinals -180 .5 units (wind gentle blowing out
Giants +142 4 units
Detroit -125 2 units (under Braden - Verlander)
DET/OAK under7 +105 (2 units)
Dodgers -115 (1 unit)
SDG/LOS under (2 units)
Blue Jays +118 2 unit wind pointed out.

10-15 -48.24
Excuse me while I smash my head on the wall for a bit.
pending ...

[therber2]

No idea, it's your graph.

A bug, maybe?



J/K

What's the method you and Therber have been using? Taking high % per inning teams?

~ VV

I think the problem here at least for me is that I am tweaking a bit too much as I go. The concept for me is that you won't do well following the same simple system over and over. I like your system Vinny, but it is that element that worries me. My biggest suggestion for you (and take it from me because I my credibility is through the roof right now ) is to figure out what filters you need to use. If you are looking at the team with the better record in 8 days I would stay away from streaks all together, and go for the teams in the 3,4,5,6 wins/losses range. A good situation might be to play against the "worse team" that is borderline win streak vs. the "better" team that is needing a win. It is hard for me to help make a system more than that; I wouldn't ever bet something completely automated, but then I just lost more units than everyone here combined ::tears::

Hope you guys did alright though.

Anyways, yes one of the things that I look for is taking a team showing better run production per inning and less runs per inning allowed than the other. After today though, just sorta feeling like I've been looking for the win in all the wrong places.::sigh::

[therber2]

5/20

Early leans

Dodgers -160 (2 units)
Yankees -140 (1 unit)
Phillies -120 (1 units)
PHI RL +150 (2 units)
PHI +1.5 -220 (2 units)
Brewers -115 (1 unit)
Rockies -140 (1 unit)
Blue Jays -120 (6 unit)
Twins +130 (4 units)
CWS -120 (4 units)
SFO -180 (10 units)
Rangers -140 (3 units)
Rangers RL +140 (3 units)
FLA +190 (2 units)
Nationals -115 (1 unit)
Oakland A's +102 (1 unit)
Braves -150 (.5 units)
Indians -125 (.5 units)

[Bigpete2010]

thanx

[Vinny Vidivicci]

Keep the faith - I spent the night breaking down every W-advantage Home and Away, incrementally.

The smallest Away advantage is a 2-1 and it ranges through 7-5 (and everything in between).

Likewise for Home advantage the range is from 2-1 up to 8-4. It's very confusing to compile, because there are so many ways to look at it. It makes me think of the garbage in-garbage out axiom. Yet the overall numbers seem to support some value.

If you or Sycoog want a copy of this breakdown, I can e-mail it or something.

Going to hit the hay - it looks like a 5-6 day for me; the 8 games we agreed on went 3-5.

~ VV

[therber2]

Keep the faith - I spent the night breaking down every W-advantage Home and Away, incrementally.

The smallest Away advantage is a 2-1 and it ranges through 7-5 (and everything in between).

Likewise for Home advantage the range is from 2-1 up to 8-4. It's very confusing to compile, because there are so many ways to look at it. It makes me think of the garbage in-garbage out axiom. Yet the overall numbers seem to support some value.

If you or Sycoog want a copy of this breakdown, I can e-mail it or something.

Going to hit the hay - it looks like a 5-6 day for me; the 8 games we agreed on went 3-5.

~ VV

Thanks Vinny, we'll get em back. Keep pluggin the numbers, but keep in mind stuff I mentioned before. If you backtrack the stuff you are looking at you have to play through the cycles of MLB and take a loss period or adjust your system around that. I'd be happy to take a look at what you have if you want me to. Email em.

[Vinny Vidivicci]

Wednesday, May 19th was 5-6, -2.7 Units (YTD : 22-14, +3.1 Units)

Thursday, May 20 -

Atlanta HF (1)
Philly HF (2)
Milwaukee AF (1)
Washington HF (5)
Houston HD (1)
San Fran. AF (2)
San Diego AD (2)
Oakland HD (3)
Toronto AF (4)
Boston HF (2)
Texas HF (5)

Just a heads-up ---------> 6 early games today

~VV

[Vinny Vidivicci]

E-mailing in a few minutes.

[therber2]

Early leans

Dodgers -160 (2 units)
Yankees -140 (1 unit)
Phillies -120 (1 units)
PHI RL +150 (2 units)
PHI +1.5 -220 (2 units)
Brewers -115 (1 unit)
Rockies -140 (1 unit)
Blue Jays -120 (6 unit)
Twins +130 (4 units)
CWS -120 (4 units)
SFO -180 (10 units)
Rangers -140 (3 units)
Rangers RL +140 (3 units)
FLA +190 (2 units)
Nationals -115 (1 unit)
Oakland A's +102 (1 unit)
Braves -150 (.5 units)
Indians -125 (.5 units)

8-10 -25.35 units

[Vinny Vidivicci]

To Sycoog, Therber, Jolmscheid, PeterPan, et al -

I spent the better part of the afternoon crunching numbers on my system and analyzing the hell out of it. The ORIGINAL PREMISE of this system was 5 out of 8. TAKE a DOG (home or away) who has WON (at least) 5 of their last 8 games (with respect to whether they are Home or Away tonight). Fade a FAVE (home or away) who has LOST (at least) 5 of their last 8 games (with respect to whether they are Home or Away tonight). Re-read what I have typed here - these are ALL DOG PLAYS.

The system then "morphed" (more or less) into a take ANYONE (except Away Faves, but I'll talk about them later), who had more wins then their opponent (with respect to whether they are Home or Away tonight - you guys get the idea by now; I'm not going to keep typing this). The results haven't been all bad. But (A special thanks to Sycoog for his programs and graphs), the records back to 2007 prove the "morphed" system is less than "earth-shaking".

I have number-crunched this system and broken it down into 8 distinct categories. The first 6 categories address all of the ORIGINAL PREMISE of the system. The last 2 categories address Home Faves and Away Faves.

Category 1 (AD +) TAKE an Away Dog who has WON 5 (OR MORE) of their Last 8.
14-13 Moneyline ..................................... 17-10 +1.5 Runs

Category 2 (AF -) FADE an Away Fave who has LOST 5 (OR MORE) of their Last 8.
12-8 Moneyline ...................................... 15-5 +1.5 Runs

Category 3 (HD +) TAKE a Home Dog who has WON 5 (OR MORE) of their Last 8.
14-13 Moneyline ...................................... 16-11 +1.5 Runs

Category 4 (HF -) FADE a Home Fave who has LOST 5 (OR MORE) of their Last 8.
9-13 Moneyline ...................................... 14-8 +1.5 Runs

Category 5 ("X") This is a Category 1 AND a Category 4 COMBINED (not any good).
4-5 Moneyline ........................................ 5-4 +1.5 Runs

Category 6 ("Y") This is a Category 2 AND a Category 3 COMBINED.
5-9 Moneyline ........................................ 9-5 +1.5 Runs

NOTE - You are playing an AD in Categories 1,4, and 5.
NOTE - You are playing a HD in Categories 2,3, and 6.

The last 2 categories are the "morphed" add-ons regarding HF's and AF's.

Category 7 (AF's) Broken down by Moneyline odds.

-150 and less 15-11 Moneyline ..................... 13-13 -1.5 Runs
-150 and more 1-5 Moneyline ........................ 1-5 -1.5 Runs (maybe a FADE here)

Category 8 (HF's) Broken down by Moneyline odds.

-140 and less 32-14 Moneyline ..................... 23-23 -1.5 Runs
-141 to -250 50-31 Moneyline ........................ 34-47 -1.5 Runs
-251 and more 3-5 Moneyline ...........................3-5 -1.5 Runs
(I might FADE a BIG HF on the Moneyline)


All "Playable" areas (in my opinion) I have highlighted in RED.

Good Luck -

~ Vinny Vidivicci

[therber2]

Very well done sir!

[therber2]

Okay Vinny, I made a note to self to start playing against LAA based on the losses due theory. Did you get my email about using innings with runs scored/innings played and the opposite for possibly a better approach to the pythagorean expectation formula? (obviously what I've been doing has been very wrong). This teams fits Catagory 1 so let's put it to the test my man. 1 unit on the ML and 1 unit on the +RL for LAA. LAA is a hot team, but what I was thinking is they are too hot for what they are worth statistically speaking.

Thing they have going for them here is playing against the Cardies piss poor offense which favors a bet of +1.5 (possible low score). Something that always tilts the scale on whether or not to play the +1.5 is my determination that the score will either be very low or it will be a tight game. I think that the Cards are false dogs.

Another thing that makes me think that there will be a low score/tight game is the SP's BB/K ratio (credit to member Solobass here who talked about this in his "press conference" at a place that cannot be mentioned ). Solo, tough break my man for the both of us; who'd a thunk that the slickster would fudge things in that one inning. No one could call that other than a law of averages concern...

We've got Pineiro against Penny.
Pineiro is 3.714 ERA in all starts with a 1.314 WHIP
in his last three games he is hot with a .84 ERA and a 1.031 WHIP
on the other hand we have Penny for STL who is 2.73 ERA and 1.272 WHIP in all starts, but
cold in the past three games with and ERA of 5.00 and 1.44 WHIP.
Pineiro's BBs/Ks ratio is a nice 1.3:4.1 all starts and 1.7:5.7 widening the gap nicely in his last three starts
Penny's BBs/Ks ratio is 1.1:4.3 all starts and also widens the gap to .3:5.3 nicely
should be a good pitcher dual, not sure who I'd lean more towards, but fact is the game will go low I think or it will be tight. Cardinals again have a bad offense while LAA is hot. I'd say overall STL is more likely to win this one at home; however, +1.5 is not too bad of a bet.

Should be some more games but this will be my only experimental play for today due to some time issues. Thank you Vinny and sycoogtit for your excellent work. Major kudos to you two.

[therber2]

A quick note to shed some insight on my current idea for a tweak on the pythagorean expectation formula. Most of you should be familiar, but if not, the formula is:

Win% = 1 / (1+ (Runs allowed/Runs scored)^2))

The tweak that I am working on is replacing runs allowed and runs scored with #innings where at least one run is allowed/innings played and
#innings where at least on run is scored/innings played

So the experimental pythag equation is:

Win% = 1 / (1+ (RAinn/RSinn)^2))

This gives slightly different results that in my opinion might be a better approach to predicting where a team should be because it factors out fluke innings where a team score 5 runs or more.

So right now as I mentioned before, if you run this equation you will see that LAA is playing far over where they should be. This is my only concern here. So Vinny this is where I see a clash with your system; perhaps you could apply this as a filter; however, keep in mind that these filters may then only yield very very few plays. I suppose what you would want is a situation where you have a home dog scoring at least 5 runs out of 8 in the process of making up for their runs due.



BOL today guys. In the hole right now, but fortunately I didn't lose a whole lot in the big picture. Just very frustrated. I know I know, I am a fool to make bets on these unproven theories. So be it.

Take care all, and have an excellent weekend.

[Vinny Vidivicci]

Thurber -

I have both the Angels and St. Louis at 4-4 of their last 8, which make neither of them a "morphed" (or any kind) of play. I just re-checked it.

I am familiar with Pythag predicting; my formula is different though. I think your formula above and mine are the same, just written differently. I use basic A-squared plus B-squared equals C-squared.

I've been playing with some numbers on the Expected Win% spreadsheet you sent me.

For the sake of record-keeping ...

Thursday was 3-5 (with AFs excluded) making the record 25-19 and probably - small juice unit-wise (since we began).

Here's Friday's plays -

Atlanta AF(3)
Cincy AF(2)
Philly HF(3)
Wash HF(3)
Mets HD(2)
Tampa Bay AF(2)
Texas HF(5)
Florida AF(2)
Minny HF(1)
Toronto AD(2)
Oakland HD(2)
Dodgers HF(4)
San Diego AD(4)

Good Luck -

~VV

[therber2]

"

[therber2]

"

[therber2]

Vinny, in place of the normal Win% = 1 / (1+ (Runs allowed/Runs scored)^2)) I am using Season to date innings with run scored/inningsplayed and innings with runs allowed/innings played in place of normal RA/RS. You can find this here: http://www.cleanuphitter.com/mlb/sta...ng_scoring.php

[Vinny Vidivicci]

Therber -

Let's take a step back.

It's not the Angels last 8 games - it's the Angels last 8 games AWAY. Likewise, it's not the Cards last 8 games - it's THEIR last 8 games at HOME. It's a site match-up premise; where you play Hot teams wherever they've been Hot; and fade Cold teams wherever they've been Cold. Sorry, but it's extremely difficult to explain this idea. I thought up till now that my explanations didn't leave anyone in the dark. Now that you brought this up, I'm wondering how Sycoog took my explanation.

Not to beat it to death - BUT to be crystal clear - let's use the Angels @ St. Louis as the example -

Angels on ROAD May 17-20; they won 2 and lost 2.
Prior to May 17th, they were HOME for 6 games from May 10-16 WE IGNORE THESE
Prior to May 10th, we go back 4 results to May 6th; they won 2 and lost 2; losing the last game in Boston and winning 2 of 3 in Seattle. This makes them 4-4 their last 8 games Against ANYONE Away.

Do the same for St. Louis' last 8 games AT HOME (Against Anyone); and they, too, were 4-4.

Note to self - How did Sycoog interpret my explanation?

~ VV

[SportsAddict]

like the thread.
keep it up!

[therber2]

Therber -

Let's take a step back.

It's not the Angels last 8 games - it's the Angels last 8 games AWAY. Likewise, it's not the Cards last 8 games - it's THEIR last 8 games at HOME. It's a site match-up premise; where you play Hot teams wherever they've been Hot; and fade Cold teams wherever they've been Cold. Sorry, but it's extremely difficult to explain this idea. I thought up till now that my explanations didn't leave anyone in the dark. Now that you brought this up, I'm wondering how Sycoog took my explanation.

Not to beat it to death - BUT to be crystal clear - let's use the Angels @ St. Louis as the example -

Angels on ROAD May 17-20; they won 2 and lost 2.
Prior to May 17th, they were HOME for 6 games from May 10-16 WE IGNORE THESE
Prior to May 10th, we go back 4 results to May 6th; they won 2 and lost 2; losing the last game in Boston and winning 2 of 3 in Seattle. This makes them 4-4 their last 8 games Against ANYONE Away.

Do the same for St. Louis' last 8 games AT HOME (Against Anyone); and they, too, were 4-4.

Note to self - How did Sycoog interpret my explanation?

~ VV

WOW. That really changes things for me, and I have to say that I am much more into the system now. I guess that you can ignore my advice that I gave as I was interpretting things all wrong. Man I feel thick headed. Vinny you are the man. I really hope that you prosper with this my friend.

I guess I have one more question that you may or may not know the answer to: do you know why this works?

[Vinny Vidivicci]

Based on the analysis from post #678, I made 5 wagers today -

Mets +1.5 RL -107 ............... Lost 1-2 .................... Won 1 Unit
Philly -Pick ML -140 ............. Won 5-1 ..................... Won 1 Unit
Houston +1.5 RL +103 ........ Won 2-1 ..................... Won 1 Unit
Toronto +1.5 RL -140 .......... Pending ...................... Losing 2-7
Oakland +1.5 RL -148 ......... Pending ...................... winning 3-0

I'll have tomorrow's up by late AM.

~ VV

[Vinny Vidivicci]

Thurber -

At THIS point, I don't even know IF it works !!
Glad you understand it though, it IS a little deeper than just going back for each team's last 8 anywhere. Heck, every newspaper in the country goes back the last 10.

How do you think Sycoog interpreted it?

Just rambling ... In one of the posts I mentioned something about a median of a "moving window" for each team. In our example above with the Angels, an AWAY "snapshot" as of Today is a 4. Tomorrow, we would discard their result at Boston (a Loss), and replace it with their result tonight at St Louis (also looking like a Loss). So their AWAY "snapshot" for tomorrow will still be a 4. It may creep up to a 5, a 6 or even a 7 or an 8, or it can slip to a 3, a 2 or a 1 or even a ZERO. If you take a median of all of these "snapshots", you will develop a power rating on a scale of 0 to 8 for the Angels AWAY. This is where we can apply your theory that you e-mailed to me. I recall Peterpan mentioning regression to the mean or something. This can be applied when a Team either over-achieves or under-achieves compared to their year-long MEDIAN of their "snapshots". Literally by taking a MEDIAN of these "snapshots" for each team will result in an Away and a Home Power Rating for all 30 teams; or a total of 60 distinct Power Ratings.

The first post I made on this thread discussing this system was late AM on Sunday, May 16th; so it will be sometime on that day.

~ VV

[Vinny Vidivicci]

Thurber -

I just sent you an e-mail with an attached spreadsheet that I use to compile these numbers.

~ VV

[Vinny Vidivicci]

Toronto needs 1 more run for me to sweep the board.

[Vinny Vidivicci]

2 out, top of 9th man on 2nd.

[therber2]

Nice thanks Vinny. I'll look into a less time consuming way to accomplish this for you. Let you know if I figure anything out. Anyone snooping around here know a source or have a thought?

[Vinny Vidivicci]

Toronto 6 runs scored on 6 solo HRs - I deserve to win this game.

[Vinny Vidivicci]

The only thought I have is to make the spreadsheet vertical instead of horizontal for each team.

Take a look at the streaks on the Totals too. Maybe we can glean something out of them too.

~ VV

[Vinny Vidivicci]

Toronto lost 8-6. Tough loss, when all 6 of their runs scored were from 6 solo HRs.

Oakland is up 6-1 in the 8th.

Looks like a 4-1 night.

[Vinny Vidivicci]

12:34 AM here, going to bed, will check the thread first thing in the AM.

~ VV

[therber2]

LOS -145 home (6) Tigers away (4)
ATL -144 away (5) Pirates home (3)
WAS -139 home (5) Orioles away (2)
STL -140 home (4) Angels away (4)
SD -120 away (6) Mariners away (2)
SFO -110 away (3) Oakland (2)

Here ya go Vinny. To save you some time. I personally like all of these plays, and it looks like SFO is the only fade; however they are barely a favorite there...

other plays:

KC +115 home (3) Col - away (3) FADE
CWS -120 home (2) FADE FLA + away (4)
TEX -150 home (7) Cubs + home (4)
MIN -128 home (5) MIL + away (4)
HOU +135 home (3) TAM - away (5)
CLE -110 home (1) FADE CIN + away (4)
PHI -140 home (5) BOS + away (2)
NYM +150 home (4) NYY - away (4)
ARI -130 home (2) FADE TOR + (4)

FINAL:

Twins -128 (3 units)
MIL/MIN o9 (.5 units)
KC +1.5 -150 (1 unit)
COL/KC u9.5 (1 units)
FLA +1.5 -177 (.5 units)
FLA +116 (.5 units)
CWS/FLA u8.5 (1.5 unit)
CIN +1.5 -150 (1 unit)
CIN/CLE u8.5 (.5 unit)
Dodgers -145 (1.5 unit)
DET/LOS o8 (2 units)
SFO -110 (1 units)
SFO/OAK o7 (3 units)
ATL -141 (3 units)
ATL/PIT u9 (.5 units)
WAS -140 (.5 unit)
Padres -108 (3 units)
CUB/TEX u9.5 (.5 unit)

[therber2]

Toronto lost 8-6. Tough loss, when all 6 of their runs scored were from 6 solo HRs.

Oakland is up 6-1 in the 8th.

Looks like a 4-1 night.

Bravo my good man