The Big Card

[darkenergy]

I was joking earlier, only played UNDER Mets .5u and loss.

[sycoogtit]

Good. I was stressing out for you.

[darkenergy]

Leans Key Heavy: abs(diff) >= 7
Mid: 3 <= abs(diff) < 7
Light: 1 <= abs(diff) < 3
Tiny: abs(diff) < 1

[darkenergy]

Good. I was stressing out for you.

Thanks, I'm gonna have to double check those plays myself (that how I find out some oddity this morning ) before I play it big. Even the data spit out an OVER for heavy play, but if a team scored 15+ runs in a day before, I probably gonna pass the play anyway.

[sycoogtit]

...

[therber2]

Thanks sycoogtit. Todays' initial leans help me. We'll see what record the leans get for today. Actually I could make a card for what those would have been by adding or subtracting the units from the card with errors. This gets us back on track. Thanks so much for doing that. Is the scraper coming along?

[therber2]

Therber -

I'm just brainstorming here -

Compute each individual team's median runs scored - WHEN THEY SCORE. Also do this individually for each team's median runs allowed - WHEN THEY ALLOW RUNS. Use Medians, NOT an Average; Medians are more predictable in nature. High run amounts skew the numbers using averages. Some teams score 1 run on average an inning, other teams score in bunches. By doing what I suggest below it will account for these fluctuations.

Multiply this by each teams Inning's-scored percentage; in your example above. In today's case, Cincinnati is 17.2 and the Mets are 13.8. Multiply this by 9 (innings) to get each team's predicted runs scored per game.

You can do this again using each team's Median runs allowed and run's allowed percentages.

You can even break it down Home and Away for each team.

You would then have an individual predicted score for each team, and a total for the over and under.

On games where the home team is the predicted winner, you can back out 1/9th of the home team's runs, because if they are ahead after 8.5 innings, they won't be batting in the 9th.

What do you think?

- Vinny

I think that you should track it.

This part I am interested in but don't fully understand what you mean: "Multiply this by each teams Inning's-scored percentage; in your example above. In today's case, Cincinnati is 17.2 and the Mets are 13.8. Multiply this by 9 (innings) to get each team's predicted runs scored per game."

Could you explain that differently?

Anyways, what I do to avoid skewed data is look at multiple sets by duration. Your median theory is a good idea, and you are free to post your card here keeping a record. It would be very interesting.

[therber2]

How many units these would have been (going off dev. from post #291)

For trackpurposes only 5/5 late card (won't be in the record; calm down): I want to see what today's card would have been. Just adding and subtracting previous deviations from the card with errors.

for this post "units" = what would have been/should have been

League Avg: 27.3

Indians (21.3) v Blue Jays (32.1): 26.7 (-0.60 differential) this would have been a NO PLAY
Athletics (26.7) v Rangers (29.3): 28 (0.70 differential) over 8.5 (.5 units)
Dodgers (28.1) v Brewers (25.9): 27 (-0.30 differential) under 9 (.5 units)
Giants (25.9) v Marlins (27.4): 26.65 (-0.65 differential) over8.5 (.5 units)
Cubs (34.6) v Pirates (18.1): 26.35 (-0.95 differential) under9 (1 unit)
Cardinals (27.7) v Phillies (27.0): 27.35 (0.05 differential) under10 (1.5 units)
White Sox (29.5) v Royals (24.5): 27 (-0.30 differential) under9 (2 units)
Reds (28.4) v Mets (28.0): 28.2 (0.90 differential) over9 (2.5 units)
Astros (23.7) v D-backs (27.8): 25.75 (-1.55 differential) over8 (2.5 units)
Mariners (21.3) v Rays (32.1): 26.7 (-0.60 differential) under 6.5 (2.5 units)
Nationals (26.7) v Braves (22.3): 24.5 (-2.80 differential) under8 (3 units)
Twins (33.6) v Tigers (26.7): 30.15 (2.85 differential) over10 (3.5 units)
Rockies (27.1) v Padres (26.7): 26.9 (-0.40 differential) under 7.5 (3.5 units)
Red Sox (33.6) v Angels (24.6): 29.1 (1.80 differential) over 9.5 (4.5 units)
Yankees (34.6) v Orioles (25.0): 29.8 (2.50 differential) over 9.5 (5 units)

[therber2]

5/6 Matchups

Totals

Bal 23.6/min 34.2
tor 29/cws 29.8
laa 27.2/bos 28.9
kc 28.5/tex 26.6
stl 23.6/phi 28.7
atl 22.4/was 28.9
tor 29/cws 29.8
cub 33.5/pit 21.4
bal 23.6/min 34.2
sfo 29/fla 31.7
tam 31.2/sea 20.9
ari 27.9/hou 21.1
mil 30.1/los 30

%Difference 21 day avg. 27.6%


12.1% cub 33.5/pit 21.4 will pit sweep Cubs?
10.6% Bal 23.6/min 34.2
10.3% tam 31.2/sea 20.9
6.8% ari 27.9/hou 21.1
6.5% atl 22.4/was 28.9
5.1% stl 23.6/phi 28.7
2.7% sfo 29/fla 31.7
1.9% kc 28.5/tex 26.6
1.7% laa 27.2/bos 28.9
0.8% tor 29/cws 29.8
0.8% tor 29/cws 29.8
0.1% mil 30.1/los 30

[sycoogtit]

Is the scraper coming along?

The scraper made that post dude! It's set to do it again tomorrow morning around 4 or 5am MST. Let me know if you want any changes before then and I'll see if I can get them in.

[therber2]

The scraper made that post dude! It's set to do it again tomorrow morning around 4 or 5am MST. Let me know if you want any changes before then and I'll see if I can get them in.

Nice. Could you email me the program?Changes for the program (unless you already thought of this) would be being able to scrape data from various durations (ie. 3day, 7 day, and 15 day). The other thing is picking sides in a matchup. I made the thread about to show how this is done. I think maybe two separate scrapers would be simpler.

[sycoogtit]

Well it's a perl script so I'll just give you access to my server. I'll e-mail you details.

[therber2]

Sounds great.

[sycoogtit]

DISCLAIMER: This is therber2's system. I'm just collecting the data for him so he can make us more money by spending his time on something else. If you're going to give points/credit to anyone, give them to him. For an explanation of what the info below means, go here.

This data has been automatically generated and posted. Please let me know if you find errors.


League Avg: 27.6
Dodgers (28.7) v Brewers (26.6): 27.65 (0.05 differential) Tiny
Twins (33.8) v Orioles (24.3): 29.05 (1.45 differential) Light
Red Sox (35.3) v Angels (24.4): 29.85 (2.25 differential) Light
Royals (25.2) v Rangers (27.1): 26.15 (-1.45 differential) Light
Giants (24.4) v Marlins (28.1): 26.25 (-1.35 differential) Light
Astros (23.1) v D-backs (28.6): 25.85 (-1.75 differential) Light
Mariners (19.7) v Rays (32.1): 25.9 (-1.70 differential) Light
White Sox (31.3) v Blue Jays (30.3): 30.8 (3.20 differential) Mid
Cardinals (25.4) v Phillies (29.4): 27.4 (-0.20 differential) Tiny
Cubs (34.6) v Pirates (19.6): 27.1 (-0.50 differential) Tiny
Nationals (27.2) v Braves (23.7): 25.45 (-2.15 differential) Light


Leans Key
Heavy: abs(diff) >= 7
Mid: 3 <= abs(diff) < 7
Light: 1 <= abs(diff) < 3
Tiny: abs(diff) < 1

[therber2]

[quote=therber2;4402649]

CUBS -1.5 RL +105 (5 units)
Cubs -152 (2 units)
Toronto +110 (2 units)

1-2 -5.84 units

wondering if I should take the Cubs again..

[therber2]

Sides 15 and 21 day

[quote=therber2;4414629]
%Difference 21 day avg. 27.6%

12.1% cub 33.5/pit 21.4 will pit sweep Cubs?
10.6% Bal 23.6/min 34.2
10.3% tam 31.2/sea 20.9
6.8% ari 27.9/hou 21.1
6.5% atl 22.4/was 28.9
5.1% stl 23.6/phi 28.7
2.7% sfo 29/fla 31.7
1.9% kc 28.5/tex 26.6
1.7% laa 27.2/bos 28.9
0.8% tor 29/cws 29.8
0.1% mil 30.1/los 30

15 League Avg: 27.6

12.4% Mariners (19.7) v Rays (32.1): 25.9
9.3% Twins (33.8) v Orioles (24.3): 29.05
7.5% Cubs (34.6) v Pirates (19.6): 27.1
5.45% Red Sox (35.3) v Angels (24.4): 29.85
3.7 Giants (24.4) v Marlins (28.1): 26.25
3.5% Nationals (27.2) v Braves (23.7): 25.45
3% Cardinals (25.4) v Phillies (29.4): 27.4
2.75% Astros (23.1) v D-backs (28.6): 25.85
2.1% Dodgers (28.7) v Brewers (26.6): 27.65
1.9% Royals (25.2) v Rangers (27.1): 26.15
0.5% White Sox (31.3) v Blue Jays (30.3): 30.8

[therber2]

Big Card 5/6

Minnesota Twins -1.5 -110 (9 units)
Rays -149 (11 units)
Red Sox -124 (5 units)
Marlins -125 (1.5 units)
Cubs -150 (10 units)
Royals +135 (1 unit) take note. If a close one; look for Rangers to strike tomorrow. This line has great value for the Royals
White Sox/Toronto *no play but take note of game today for possible play on CWS tomorrow.

Phillies RL -1.5 EV (2 units)
Cardinals -1.5 +335 (1 unit) risk
**As long as the game doesn't end up close you make a profit here. Note to self. System for predicting games that won't be close with expanded odds?

Brewer +113 (1 unit)
PHI/STL under8 (.5 units)

[therber2]

Test

JUST A TEST

7 days

League Avg: 27.5
Dodgers (30.0) v Brewers (16.2): 23.1 (-4.40 differential) Mid
Twins (33.3) v Orioles (23.4): 28.35 (0.85 differential) Tiny
Red Sox (37.1) v Angels (22.2): 29.65 (2.15 differential) Light
Royals (18.5) v Rangers (23.5): 21 (-6.50 differential) Mid
Giants (30.8) v Marlins (26.6): 28.7 (1.20 differential) Light
Astros (11.1) v D-backs (28.6): 19.85 (-7.65 differential) Heavy
Mariners (17.6) v Rays (32.3): 24.95 (-2.55 differential) Light
White Sox (32.8) v Blue Jays (40.0): 36.4 (8.90 differential) Heavy
Cardinals (26.2) v Phillies (28.6): 27.4 (-0.10 differential) Tiny
Cubs (35.0) v Pirates (22.7): 28.85 (1.35 differential) Light
Nationals (33.3) v Braves (32.8): 33.05 (5.55 differential) Mid

This.is.a.m.a.z.i.n.g. I am going to look through durations to find some patterns. Obvious pattern of the day would be the Cubs. Their % went like this:
21day: 33.3%
15day: 34.6%
12day: 36.8%
7day: 35%
3day: 26.9%

While the Pirates did:
21day: 21.8
15day: 19.6% dead last
12day: 23.4%
7day: 22.7%
3day: 28%

Eat this up with a spoon boys. Who is hot and who is not. I made the wrong call today; lesson learned? I think so. Keep an eye out for teams dead last and now hot.

[sycoogtit]

Holy crap, 100 big ones! Thanks senor, I'm glad you're smart enough to find stuff like the Pirates/Cubs that you posted above.

[therber2]

Holy crap, 100 big ones! Thanks senor, I'm glad you're smart enough to find stuff like the Pirates/Cubs that you posted above.

Well, its too late for that. The deed is done. Lesson learned though. I sent you an email.

[sycoogtit]

Using therber's system and the fancilicious new way to see the data, here's what my leans might have been for today (trying to forget what the scores were):

Dodgers v Brewers
Dodgers have been in a pretty bad slump the last couple games while the Brewers have been picking it up pretty drastically. Brewers have a 16.6 RS% advantage over the last couple games combined, but going beyond the last two days the Dodgers dominate. Their 15-day matchup RS% is pretty much the same as the league avg, although it's under for much of the last 15 days and under 2.65 for 1-day.
ML: Slight Brewers edge .5 unit
OU: 1 unit

Twins v Orioles
Orioles RS% is trending way up over the last two games, but the Twins just dominate from the last 1 day through the last 21 days. Their matchup RS% is consistently over the league avg from 1-21 days, with it being over 1.45 for the last 15 days and over 5 for 1 day.
ML: Twins 3 units
OU: 2 units

Red Sox v Angels
Neither team is hot or cold, and the Red Sox just beat the snot out of the Angels from 1-21 days, with an average being in the teens I would guess. 15-day matchup RS% is 2.25 above avg, but -6 and -8 for 1 and 2 days. They average over, but just slightly.
ML: Red Sox 4 units
OU:

Royals v Rangers
Rangers have been spiraling downward, but the Royals don't look to good either. Rangers are 1-6 points over from 4-19 days, but the Royals are 11 points over for 1 and 2 days. Their matchup % is under the league avg across the board, being -1.45 for 15-day.
ML: Royals 2 units
OU: 2 units

Giants v Marlins
Both teams seem a little cold, with either team taking the advantage from 1-8 days but the Marlins taking the edge from 9-21, with a 3.7 point edge for 15 days. They both appear to be streaky matchup-wise. From 13-17 days they're slightly under, but slightly over around that chunk.
ML: Marlins 1 unit
OU: 1 unit

Astros v D-backs
Astros have been sucking the past 2-11 days, but jumped 10 points yesterday. The D-backs haven't been too hot either, but they consistently own the 'stros with a 10+ point advantage from the last 3-10 days, and a 5.5 point advantage for 15 days. These teams seem to never score because their matchup % is always way under the league avg.
ML: D-backs 2 units
OU: 4 units

Mariners v Rays
I'm getting tired, and this is all old stuff anyway. Just gonna cut to the chase because I'm interested how this system would have worked today. Maybe I'll do this for tomorrow games.
ML: Rays 3 units
OU: 1 unit push

White Sox v BJs
ML: BJ's 1 unit
OU: 4 units

Cards v Phils
ML: Phils 2 units
OU: 1 unit

Cubs v Pirates
ML: Pirates 1 unit
OU: 1 unit

Nationals v Braves
ML:
OU: 1 unit


Sweet! Keep doing what I'm doing and just fade myself and I'll be rich! I didn't get a single total right. That's gotta be a record.

[therber2]

Minnesota Twins -1.5 -110 (9 units)
Rays -149 (11 units)
Red Sox -124 (5 units)
Marlins -125 (1.5 units)
Cubs -150 (10 units)
Royals +135 (1 unit) take note. If a close one; look for Rangers to strike tomorrow. This line has great value for the Royals
White Sox/Toronto *no play but take note of game today for possible play on CWS tomorrow.

Phillies RL -1.5 EV (2 units)
Cardinals -1.5 +335 (1 unit) risk
**As long as the game doesn't end up close you make a profit here. Note to self. System for predicting games that won't be close with expanded odds?

Brewer +113 (1 unit)
PHI/STL under8 (.5 units)

3-6 -11.78 units

[sycoogtit]

Shazam therber, do you have points coming out your ears? Thanks for the generous donation, AGAIN!

[therber2]

Those are great notes sycoogtit. What was very notable about today I think was what happened with the Cubs like I mentioned before. The Rays, Red Sox, and Twins were all the best picks as far as "beating the snot" out of the other guy stats wise. Now why didn't the Twins make it. I'm not sure but the game was close. I will be taking them pretty damn big tomorrow.

[therber2]

Shazam therber, do you have points coming out your ears? Thanks for the generous donation, AGAIN!

points shmoints I'll get em back tomorrow.

[therber2]

One thing to say sycoogtit. Those stats are a starting point. For overs and unders the best way that I am seeing to do things is to put more focus on the 15-18 day duration, and then more recent performance with a grain of salt. From that starting point the game needs to be capped in a variety of ways. Recall the deviations (sometimes very strong) from the starting point.

7 day went 9-2 for sides

[therber2]

5/7

Buffet style 5/7

Minnesota Twins -1.5 -110 (12 units) CANCELED
NYY/BOS o9 -125 (6.5 units)
Tigers -105 (5.5 units)
Nationals +105 (5 units)
Mets -115 (5 units)
D-Backs +115 (4.5 units)
STL/PIT u8 EV (4.5 units)
Padres -120 (4 units)
LAA/SEA u7 -125 (4 units)
Rays RL -1.5 +105 (3.5 units)
SDG/HOU u9 -110 (3 units)
ATL Braves +120 (3.5 units)
CUB/CIN u9 -115 (3 units)
Cubs -110 (3 units)
KC/TEX u8 -105 (3 units)
STL -1.5 -130 (3 units)
*Pittsburgh +1.5 +105 (2 units)
BAL/MIN o8.5 +105 (2 units)
OAK/TAM o7.5 -105 (1 units)
KC +110 (1 unit)
Rockies +155 (1 unit)
NYM/SFO u7 EV (1 unit)
ARI/MIL o9.5 -110 (1 unit)
Angels +127 (.5 unit)
CWS/TOR o8.5 +110 (.5 units)
CWS -106 (2 units) Probably a better bet on Blue Jays because of recent results; however, the White Sox have Buerhle on the mound today. If I have to pick I'd say Jays is a pretty good bet.
*Blue Jays +106 (.5 units)
NYY -105 (.5 units)

[sycoogtit]

The data seems to point to Giants ML -- why are you picking Mets?

[sycoogtit]

Here's what I've got. I'm just doing 1 unit across the board.

Mariners/Angels u7 -120 1.2u tw 1
Rays/A's o7.5 +117 1 tw 1.17
Padres/Astros u8.5 -101 1.01 tw 1
Cards/Pirates u8 +106 1 tw 1.06
Red Sox/Yankees o9.5 +110 1 tw 1.1
Angels ML +123 1 tw 1.23
BJ's ML +106 1 tw 1.06
Rays ML -122 1.22 tw 1
Dodgers RL +142 1 tw 1.42
Padres ML -118 1.18 tw 1
Tigers ML -104 1.04 tw 1
D-backs ML +119 1 tw 1.19
Cubbies ML +120 1 tw 1.2

[therber2]

The data seems to point to Giants ML -- why are you picking Mets?

Ok let's dissect this a bit. Consider the graph attached.
Yes, if you had to choose it might appear that the Giants would be the ones to take from this picture. It is too close though to produce a strong enough lean either way long term; however, take a look at the end: the Mets shoot up sharply. They are a bit hotter than the Giants. Still not enough though so I take it to the streets. The Mets haven't been considered a very decent team for a while, and now they are contenders thanks to Mr. Pelfrey who is pitching today. On the other hand, let's think about what the Giants have to offer: GREAT PITCHING. Well Cain and Lincecum are taking the day off, and we've got Pelfrey in today.

What else does this game have going? **It is the end of the week, the Giants won two series and the Mets just lost their last one. The Giants are visiting New York to have a good ol time; while the Mets mean business.

...sometimes you just need to look at the game this way. The numbers didn't tell me anything other than the Mets are pretty good and now prepared for battle.

[therber2]

Here's what I've got. I'm just doing 1 unit across the board.

Mariners/Angels u7 -120 1.2u tw 1
Rays/A's o7.5 +117 1 tw 1.17
Padres/Astros u8.5 -101 1.01 tw 1
Cards/Pirates u8 +106 1 tw 1.06
Red Sox/Yankees o9.5 +110 1 tw 1.1
Angels ML +123 1 tw 1.23
BJ's ML +106 1 tw 1.06
Rays ML -122 1.22 tw 1
Dodgers RL +142 1 tw 1.42
Padres ML -118 1.18 tw 1
Tigers ML -104 1.04 tw 1
D-backs ML +119 1 tw 1.19
Cubbies ML +120 1 tw 1.2

I am glad you posted this. Let's compare where we went wrong and right. Also, I am looking to see if that 7 day trends holds up again today. The most telling duration may actually be dependant on the time of the week. Today, I would guess that 3-5 day duration is most relevant, and normal days 21 day to 15 day. On middle of the week, and a tight matchup, maybe something in between say 7. When we have things logged we can analyze this further.

[therber2]

Tell me what you think the most blowout game was at the end of the day please. Also tell me what you think was the most nailbiting close game. We'll see what the numbers were saying.

[skype 1]

bol!

[sycoogtit]

Well this Yankees game has been a real nail-biter...

You were smart enough to get both the ML and over. I only got the over.

[therber2]

buffet style 5/7

minnesota twins -1.5 -110 (12 units) canceled
nyy/bos o9 -125 (6.5 units)
tigers -105 (5.5 units) postponed
nationals +105 (5 units)
mets -115 (5 units)
d-backs +115 (4.5 units)
stl/pit u8 ev (4.5 units)
padres -120 (4 units)
laa/sea u7 -125 (4 units)
rays rl -1.5 +105 (3.5 units)
sdg/hou u9 -110 (3 units)
atl braves +120 (3.5 units)
cub/cin u9 -115 (3 units)
cubs -110 (3 units)
kc/tex u8 -105 (3 units)
stl -1.5 -130 (3 units)
*pittsburgh +1.5 +105 (2 units)
bal/min o8.5 +105 (2 units) canceled
oak/tam o7.5 -105 (1 units)
kc +110 (1 unit)
rockies +155 (1 unit)
nym/sfo u7 ev (1 unit)
ari/mil o9.5 -110 (1 unit)
angels +127 (.5 unit)
cws/tor o8.5 +110 (.5 units)
cws -106 (2 units) probably a better bet on blue jays because of recent results; however, the white sox have buerhle on the mound today. If i have to pick i'd say jays is a pretty good bet.
*blue jays +106 (.5 units)
nyy -105 (.5 units)

15-7 +16.8 units