berryhorse tracker

mrpickem

mrpickem

SDQL Badass Stat Boss
There seemed to be some interest in this guy from conversation yesterday, so I an going to track him publicly here starting today.

Dudes twitter https://twitter.com/berryhorse29
Dudes model https://goo.gl/VDYZZq

I back dated 5 days although he had no plays Friday as I was already tracking him and I can vouch they are legit. I am showing results 2 ways with his unit value and flat betting with $100 risk on all plays fav or dog. Out of the 20 lays pn this first 5 days he only took 2 dogs. Also I notice he played two games under 5%...like 4.5 or so, not sure why.

View attachment 33859

Also I added likely plays for today, but they are not official until he adds lineups.

Let me know if there is any interest in this and I will post it daily..I been tracking him since last Tuesday myself and of course there is his sheet and there seems to many followers and I've seen no one claim it was entered anything but honestly.

Also made my tracking sheet available here
 
Last edited:
value % means what? the higher the value the more he likes it or is determined in his mind that the line is way off?
 
It's easiest to read his FAQ on the same link for all questions on the model. It's well thought out.
 
Noles4life said:
value % means what? the higher the value the more he likes it or is determined in his mind that the line is way off?
Click to expand...

It's based on value over line.

Here's his quote(s)
I've built a model that assigns a probability for an MLB team winning a game given their starting pitcher and lineup, and their opponents'. This spreadsheet, my posts on the daily MLB reddit threads, and my twitter account @berryhorse29 are used to help share this with the public. Every recorded bet and record are the results for betting on teams my model gives an over 5% edge to in comparison to the implied probability of the line at Pinnacle. I don't at all claim for this to be optimal - I welcome anyone to come up with subsystems for improving this. Just because I tweet something out does not mean it has over 5% edge at YOUR sportsbooks, and just because I don't tweet something out doesn't mean you don't have over 5% value at YOUR sportsbooks. It also doesn't mean I bet everything on here myself. There's all kinds of ways to improve your returns. This isn't my full-time job. Please don't rely on me to tweet out every time there is a 5% disagreement. I've done my best and have covered ~99% of the fits, but it's possible I'll be busy and miss some.
Click to expand...

Why 5%?

The 5% threshold is frankly very arbitrary. Without a full season of data I'm uncomfortable making any claims about an optimal threshold and you can certainly convince me a higher or lower number would be better as the cut-off. To me, 5% seems like a good trade off between efficiency (winning at a high % and only betting when you're confident) and volume (betting frequently enough to actually make money and not just bet once a month).


Why do I bet so many units on each game?

I believe it makes sense to bet more money on games I am more confident in, and rather than doing "confidence brackets" it makes intuitive sense to me to just have this unit size linearly proportional to the actual level of confidence, or % edge that the model has over the vegas line. I also strongly reject the idea of betting games on a scale between 1 and 5 units. I personally tihnk it's ridiculous to be 5 times as confident in something compared to another wager that also has value. If this is the case it seems like you either shouldn't be betting the 1unit plays or should only be betting the 5 unit plays. The way I assign units between 5 and usually 10ish, but potentially up to 20 seems way more logical, but I don't claim any of these ideas to be statistically optimal over others and am planning on doing a full analysis of potential unit breakdowns after the season ends. You're welcome to use whatever units you want.
Click to expand...
 
Thanks Mr. P for the tracking, i have already been looking up that link today myself. I saw you posted that around 2 pm and just another piece of information for you, yesterday when I looked, he said that the stats wouldn’t be update or correct until like an hour or 2 before each game because he has to wait to see lineups and such. Just a friendly reminder as there can always be some changes. Not sure if you saw my post last night thanking you for introducing me to this, but just looking at yesterday alone, there were 15 total games and his model/math equation essentially called 11/15 right, even though there were only a few over the 5% threshold. That to me is pretty sick. I’m gonna follow for a few days before I make any serious plays or tails, but still. If that model can predict the better bet in each game, even if it doens’t reach the 5% threshold, that is still an awesome guide. Ill take 11/15 winners any day of the week.
 
I’m gonna roll with three of the teams listed as having a statistical edge, philly, sf, and Houston. See how it goes. I think I am also gonna go with atl even though his tracker gives the slight edge to miami
 
still just observing this guy

tonight phils are not a play because they would have to be at least +160 to be considered a play?

also it looks like Astros are becoming or have become a play
 
View attachment 33865

these are games in model tonight, first4 have final lineups in, the listed lines are open and now from pinny for the home team
appently the model figures based on home line, so he explains the phi 6.63% value is based on +151 when you really only get +141 after vig...thats why phi not a play. I took BOS earlier myself and I'm gonna let it ride
 
The stat advantage went way up for miami and of course i bet against them. Guess i hope the model is wrong there
 
Big_Degenerate said:
still just observing this guy

tonight phils are not a play because they would have to be at least +160 to be considered a play?

also it looks like Astros are becoming or have become a play
Click to expand...
I still have to look more into the whole “not a play based on the juice” classification, but essentially i think it is something close to what you said, meaning it has to do with the listed odds and how that affects the percentage
 
Stats also saying that minn and StL both have the statistical edge. I can believe on STL, but just can’t bet against Cleveland with Santana. Oh well. Guess we will see. So far, on philly and atl
 
final lineups in and no plays tonight :(

some promising plays for tomorrow..early lineups default to last game so monday will have a lot of skewed lineups until real line up entered, because of Sunday rest

View attachment 33867
 
Yesterday 6-3 based on teams listed with statistical advantage. And once again, 2/3 losses were dogs and the other was basically a pick
 
Johnnyonthespot said:
All yall $100 betters time to put down $1300 on Clayton.
Click to expand...

Think you're looking at it from the wrong angle, Johnny.

If the average bettor of $100 sees this play, what is the average amount of units this guy puts down as a bet?

That's the question I think people need to figure out before trying to tail a guy with an odd structure / model like this.

Figure that out and then you can bet in the context of his model.

Quite honestly, I have no idea what the answer to that question is though.
 
Okay, I own a financial consulting company, so I am strong with Excel...took a little closer look at his model and units wagered / units to win

So on AVERAGE...(I used every wager he's made so far this year from his spreadsheet, FYI)

Average Play for 2018 - 7.65 Units Risked
Average Play for 2018 - 9.57 Units To Win

So he's basically betting lines averaging +125 (over the course of the whole season on average).

So to take it further in Johnny's example....the $100 bettor is equivalent to 7.65 Units.

He is playing Padres tonight risking 13.44 to win 13.8432

Average $100 bettor (who is trying to tail him) should be wagering (13.44/7.65) $176 tonight on the Padres.

Wayyyyyyy geeky a math guy's understanding of how to best make his model and real world gamblers equivalent, but think that may be helpful for a couple of people reading it.

Or...I'm just too much of a nerd into the nuances of gambling/gamblers...you guys can be the judge of that. LMFAO
 
I like his number crunching for comparable advantages.

I need to break down his picks as if each game was to win one unit with his odds played and see where he really is.

People seem to be impressed on his units won which is based solely on his weighting which is ridiculous to me.

How can you benefit from his success unless you play it in the exact random way he does?

How can he justify even saying 13 units?

I'm definitely not looking to argue with you scarf but I really dont get it.
 
Last edited:
Johnnyonthespot said:
I like his number crunching for comparable advantages.

I need to break down his picks as if each game was one unit with his odds played and see where he really is.

People seem to be impressed on his units won which is based solely on his weighting which is ridiculous to me.

How can you benefit from his success unless you play it in the exact random way he does?

How can he justify even saying 13 units?

I'm definitely not looking to argue with you scarf but I really dont get it.
Click to expand...

You're not arguing at all...your point is valid in that just him posting these odd numbers seem strange as hell to the Joe Average gambler.

Hopefully the above post kind of brings his nerdiness model and real life money into a better understanding.
 
basically 5 units is his minimum play..so call that one unit...since 13 is 2.6 times 5....you can consider this wager 2,6u if like it better with lower units :haveanidea:
 
mrpickem said:
basically 5 units is his minimum play..so call that one unit...since 13 is 2.6 times 5....you can consider this wager 2,6u if like it better with lower units :haveanidea:
Click to expand...
Thanks, but I'm talking about for the season when every game is to win 1 unit. Do you know the answer?
 
Johnnyonthespot said:
Scarf what is his win percentage and units won when you normalize each wager for 1 unit. Seems like you got that handy
Click to expand...

His average bet wagered is 7.65 and it is played at an average line of +125

His record on the season is 166-138

So 7.65 x 1.25 x 166 = 1587.38 (Total Won In 2018 In This Simplified Example)

And 7.65 x -138 x $100 = -1055.70 (Total Lost In 2018 In This Simplified Example)

1587.38 - 1055.70 = 531.68 (Net Won Lost in 2018 In This Simplifed Example)

531.68 / 7.65 = +69.5 Units (Net 2018 Total Divided by Average Bet Size)

This is a super, super, super rough example...and I'm sure you can analyze it other ways, but that's approximately where I think he's at in terms of a $100 bettor.

His record of 166-138 = 54.6% for the year betting an average line of +125.
 
Johnnyonthespot said:
Thanks, but I'm talking about for the season when every game is to win 1 unit. Do you know the answer?
Click to expand...
Tough to answer that without geeking out into the numbers and this crazy ass model he has. Just trying to give some equivalencies to the real world. Just don't have an abundance of time as I'm slammed at the office, but I always find time to try to make CTG better for my fellow degenerates. :shake:
 
54-55% is a huge number compared to the avg joe, it's why I could never flat bet

No way I'd hit much more than 50% in any sport, obviously it would usually be less over a season
 
This guy gives his stuff out for free? Keep this thread going. Interested to see how he does once tracked like this.
 
Inspekdah said:
This guy gives his stuff out for free? Keep this thread going. Interested to see how he does once tracked like this.
Click to expand...

He actually on his Tweeter page tracks it all himself, Spek. Guy updates his sheet i believe daily...link to his entire spreadsheet is all there.
 
Inspekdah said:
This guy gives his stuff out for free? Keep this thread going. Interested to see how he does once tracked like this.
Click to expand...
Check out his Twitter, he pretty much lays it all out there on the pinned tweet. Like this guy's attitude...not that I'll tail but will follow and see how it goes.
 
Don't forget guys...his model is 100% based on the EXACT starting lineup for each team. If there is a scratch or the lineups aren't out yet, his plays are absolutely going to change. In his model, every single player gets a value that ends up calculating the overall value in a potential bet on each team (potentially)
 
scarf31 said:
Don't forget guys...his model is 100% based on the EXACT starting lineup for each team. If there is a scratch or the lineups aren't out yet, his plays are absolutely going to change. In his model, every single player gets a value that ends up calculating the overall value in a potential bet on each team (potentially)
Click to expand...

Yeah, he even calculates for bullpen usage and injuries to relievers

He says he is a bitcoin developer whatever that is, but he is semi retired and allegedly wealthy
 
You must log in or register to reply here.

Similar threads

Wiretowire
MLB Tuesday 10/1
Replies
12
Views
165
Paul Patrol
P
mrpickem
MLB in July ⚾
2
Replies
77
Views
2K
mrpickem
mrpickem
Wiretowire
MLB Wednesday 10/2
Replies
2
Views
112
Wiretowire
Wiretowire
VirginiaCavs
Sunday Picks Preview Article
Replies
5
Views
57
2daBank
2daBank
VirginiaCavs
Twins vs. Red Sox (Game 1) Preview Article
Replies
1
Views
48
VirginiaCavs
VirginiaCavs
Back
Top