vizIIsto's Betting Advisor

[vizIIsto]

YHE UGYS, MI' BKAC!

So what's this? Well, I'm really, REALLY into stats and numbers, and I had a little thought the other day about whether or not I could earn some cash betting on darts matches. So I've decided to start an experiment where I'm calculating what are the best bets to make according to the stats, and see if I earn a profit out of it!

You might be wondering: what 'stats' are you using? Well, I'm going by the FDI form rating. It's a rating designed by a guy named Lendel Faria which tries to measure how well someone is playing by giving them a number rating, like an ELO rating in chess. He has a twitter account called FDI Darts, and you can also find the FDI leaderboards on dartsactueel.nl/rankings/fdi-rankings/
Now, for most TV tournaments, he simulates matches going by the ratings he generated himself, which should be accurate enough, to give percentage odds of each player winning/losing their match. I'm going to compare those percentages to what the bookies are offering, and place bets accordingly to try and maximize my profit.

Here's a little (too long) basic explanation on how you can earn money off of sports betting:
A betting office's win model is based on how many bets are made. For example, in a darts match, one player is the favorite while the other is the underdog. 3 people place a bet of $1 on Player A, whereas the fourth places a similar bet on Player B. The betting agency would get $4 from its gamblers.
If Player A wins, then the three people combined should not get more than $4, otherwise the betting office doesn't make any money. So let's say they each get $1,30. That means that 10 cents would remain and go to the betting agency's profit. The odds that the betting office gives you here is 1.30, meaning you get $1,30 for every $1 you bet with. The fourth better would get, for example, $3.90 when Player B wins, because that again leaves the betting office $0,10 profit.
A betting office functions because the odds it gives you are made to almost always guarantee profit for the betting agency. That's why you'll never see 2.00 plays 2.00, because that would mean that if the odds are also evenly spread out, the betting office doesn't make any money. That's why you'll usually see 1.90 against 1.90 for what the bookies expect to be an even match.
Coming back to the example, let's say that you found some form ratings which suggest that the chance that Player A wins isn't just 75%, but 90%. If you were to place a $1 bet on Player A, you would earn $0,30 9 out of 10 times, and lose $1 1 out of 10 times, IF the form ratings are indeed reliable. If you add that up, that means you'd win $1,70 if the match was played 10 times, where A wins 9 times (+$2,70) and B only once (-$1,00). So, for every dollar you'd bet, you should expect to win $0,17, or a 17% profit.

This is how I'm calculating my average expected value, or EV.
For this experiment, I will give myself 100 points, which I can distribute across all matches per session, after which I calculate how much I get in return. I will then do my calculations for the next day, and so on. I might continue after the World Matchplay.
For the World Matchplay, I'm going by the odds of bwin.com. I wanted to reach skybet.com, but it's unavailable in my country.

Finally, I'll also keep track of how much YOU could have earned if you were to listen to my recommendations. I'm placing a $10 bet (fictionally) on the bet that I deem the best bet, and then I'll keep track of how much money you'd have won or lost.

TL;DR: I'm starting a thread where I experiment how much money I could have earned by betting on darts matches by using FDI ratings and the odds provided by sports.bwin.com, using my calculating skills to try and earn a big profit. I'll also keep track of how much you could have earned if you were to listen to my advice.

[vizIIsto]

Assuming that bwin's odds don't change over the next two days, here is my analysis of Day 1's betting:

- I started today with 100 points, but for convenience I'll use $100 from here on. I have decided to bet with 65 of them.
- Gary Anderson vs. Justin Pipe is the only match where neither of the players' winning odds are giving me a profit on average.
- Here are the other four matches, with bwin's odds and the bets that should give me a profit on average:

• Simon Whitlock (1.80) vs. Ryan Joyce (2.00): Betting on Joyce should win $1.30 on average for every $10 bet.
  Krzysztof Ratajski (1.44) vs. Jermaine Wattimena (2.75): Ratajski wins $1.32 for every $10.
  James Wade (1.33) vs. Keegan Brown (3.30): Wade wins $1.13 for every $10.
  Michael van Gerwen (1.08) vs. Brendan Dolan (8.25): Dolan wins $0.81 for every $10.

- I'm placing the following bets: $10 on Joyce, $25 on both Ratajski and Wade, and $5 on Dolan. This should give me an average profit of $7.83 for the day, or +7.83% compared to the beginning of the experiment. I would turn a loss 30% of the time, and it's very unlikely that I'll lose more than 12%. In about 45% of occurrences, I'd have earned more than my expected value.

Recommended bet: Ratajski has a high chance of winning against Wattimena (79%) and is therefore a safe bet. It also gives the highest EV for an individual bet, at +13.2%. Therefore, I recommend betting on this player.

My balance at the start of this day: 100 points / $100
Your profit/loss at the start of this day: $0

My balance after this day: ???
Your profit/loss after this day: ???

[ILAD]

Can you cheat in this?

[Safc]

Can you cheat in this?

[vizIIsto]

I mean, I could edit my posts after the matches have ended to give me more profit...

[vizIIsto]

I'll actually expand on this and not only include match odds on itself, but also outright champion odds and other things the bookies have to offer like 180 count and high finish odds, things like that. I'll include bwin's odds for every chance I decide to bet on. I'll therefore also increase my bankroll to 500, to give me more room to play with.

[Mabbsy]

Have you not finished spending your World Championship winnings yet ?

[ILAD]

Have you not finished spending your World Championship winnings yet ?

SAFC pinched it.

[Safc]

Have you not finished spending your World Championship winnings yet ?

SAFC pinched it.

Aye I used my Secret decoding device which I purchased from home bargains

[sennafan24]

Where is Nigel?

[Safc]

Where is Nigel?

He makes Millions not thousands

[SuperSwede]

I find FDI generally predicts too much in terms of both % and winning margin for the favourites. This is from casually looking mind not in depth.

I'm + darts betting over a 11/12 year period, nothing life changing but decent, i'd question why with very similar EV guesstimated your Ratajski stake is 2.5x that of your Joyce stake. Similar EV should mean similar exposure, if you're backing off because of fear of variance then it means you are staking too big a % of bank to begin with.

[bojeru]

Can I get an over/under on the day you go busto?

[vizIIsto]

I find FDI generally predicts too much in terms of both % and winning margin for the favourites. This is from casually looking mind not in depth.

I'm + darts betting over a 11/12 year period, nothing life changing but decent, i'd question why with very similar EV guesstimated your Ratajski stake is 2.5x that of your Joyce stake. Similar EV should mean similar exposure, if you're backing off because of fear of variance then it means you are staking too big a % of bank to begin with.

I understand your point of why you advise against changing bet sizes according to the percentages for bets with similar EV. I'm going by trying to minimize the amount of risk I have to take over how much I'm expected to win in the long run. For example, I could bet only on one of the four games, but since I don't have any backups for when I'm wrong, it's more risky and I'm avoiding that. Besides, if the FDI's percentages which you state are inaccurate are somewhat off, I am still more likely to have a backup somewhere if I place more bets. In other words, I'm trying to make the chance that I lose a lot more than I'm trying to maximize my winnings.

And about FDI: One strong argument against the rating is that players with high ratings are bound to lose their form at some time, which would suggest it's better to go by, let's say the 3-year running average FDI rating of each player. If the current rating is lower than their average, this argument suggests it's actually very likely they will regain form sooner than those whose average is lower than their current rating. If this is true, the percentages given by the FDI ratings should be far inaccurate.
On that note, even if that's true, FDI's estimations would still need to be far off to turn an expected profit into a loss on the long run. For example, if all of the percentages of odds, that I'm going to place bets on in the next post, would actually be 30% smaller in reality compared to what FDI suggests, I'd still be able to expect to have a profit of over 20% on average, also without changing the bet sizes (I'd be placing bets on some odds that are no longer profitable). So if the FDI percentages are at least somewhat accurate, I think I should be able to end on the positive side of the balance in the long run.

Can I get an over/under on the day you go busto?

Unfortunately I'm no betting company myself, but I'd like to accept the challenge. Since I'm going to try to manage my bankroll as efficient and safe as possible, I think the odds of me going bankrupt at all are very small, but if you wanna 'gamble' evens/evens of me going bankrupt either before or after day 8 (that's the semi's) you are free to do so.

[vizIIsto]

So the tournament starts tomorrow, and because of that I have enough time to consider placing bets on odds that cover the entire tournament. Outright odds of players winning, odds of players getting eliminated at a certain stage or odds of players reaching a certain stage. Bets that need to be made before the tournament kicks off. So I'll cover those today.

I have decided to wager about half of my bankroll on these bets. I'm keeping 256 points in my back pocket (since it's nicely dividable by 2 multiple times and that means I can bet in such a way I'll rarely ever lose a big portion of my bankroll early on), and the remaining 244 are spent on pre-tournament bets. Here are my bets, with Betfred's odds for each of them (yes, I switched from bwin to WM's sponsor Betfred), including FDI's suggested percentage of success and the EV for all of the odds.

Quarter Bets:

• Ratajski winning his quarter of the draw (4/1): I can expect to win $8.5 for every $10 I place on this bet on average | 36.9% success rate.
  Bet size: 42 points
  Wright winning his quarter (13/8): +$4.7 per $10 placed | 56.1% success rate.
  Bet size: 35 points
  White winning his quarter (6/1): +$4.5 per $10 placed | 20.9% success rate.
  Bet size: 13 points

Specials Bets:

• One of the Big 3 (MvG,Wright,Price) winning the title (8/11): +$2.3 per $10 placed | 71.1% success rate.
  Bet size: 22 points
  Ratajski reaching the Final (14/1): +$11.3 per $10 placed | 14.2% success rate.
  Bet size: 22 points
  Wade reaching the Final (16/1): +$13.3 per $10 placed | 13.7% success rate.
  Bet size: 24 points
  Anderson getting eliminated at Round 2 (9/4): +$4.5 per $10 placed | 44.5% success rate.
  Bet size: 27 points
  Aspinall getting eliminated at Round 2 (10/3): +$4.9 per $10 placed | 34.4% success rate.
  Bet size: 23 points
  Wright vs Ratajski in the Final (50/1): +$16.7 per $10 placed | 5.2% success rate.
  Bet size: 12 points

Champion Bets:

• Wright winning the Title (13/2): +$7.7 per $10 placed | 23.6% success rate.
  Bet size: 24 points

You are probably wondering: How did you decide how many points to place on each bet? Well, I have my own formula for that, but I won't share nor explain it here. I do want to say that the reason I'm not placing massive bets on odds that are giving me insane average payouts is for one particular reason: risk management. Although I'd win more in the long run, I am aiming to have an as consistent as possible profit, if any. I could place all my points on the ones that should double my wagers on average according to the stats I'm working with, but since they are very risky, the chance that I'm losing a lot of my input is much larger than it is now. In fact, I'm expected to make a loss on these bets combined about 35.5% of the time, and in a bunch of those occasions, I'm losing a very small amount.
My EV for these bets is 181, which means I'm expected to make a 181 point profit on average according to my calculations, a.k.a. an average expected profit of 74% of my bets / 36% of my current bankroll (including the bets that are running now).

My recommendations:
- According to FDI's estimations, Ratajski is actually the most likely player to reach the Semis for his portion of the draw. For his 36.9% success rate, the expected return is also very high, standing at 85% profit. This is the bet I would place money on if I could only choose one.
- If you like to bet with high risk, high reward, a.k.a. the SuperSwede system, betting on Wade reaching the final seems like a decent choice, since you're more than doubling your wager on average, as you get back 17 times your placed bet if the 13.7% estimated chance occurs.
- If you like to bet with more safety, then placing a bet on Wright winning his quarter of the draw is my advice. You're more likely to be right than wrong, according to FDI's percentages, but nonetheless Betfred is giving you back more than twice the amount of money you betted with.

Remember how I said that I was going to keep track of how much winnings you could've made by placing set amounts of money on my recommended odds? Here's how I'm going to do that. I'm going to virtually place a $10 bet on my top recommendation, which is Ratajski reaching the Semi-Finals, and two $5 bets on both the riskier and safer bets.

About this, here are our current bankrolls:
My virtual balance before this tournament: 256 points (excluding: 244 placed on pre-tournament bets)
"Your" virtual profit/loss before this tournament: -$20

Happy betting!

[SuperSwede]

"If you like to bet with high risk, high reward, a.k.a. the SuperSwede system"

That is the opposite of what I do and suggested, I am the most cautious bordering on boring "gambler" in history but made a living off realising EV in another field for almost 15 years.

I was't suggesting you raised all your stakes to the Rataski level, I was suggesting your Ratajski level was too high and you were over exposing yourself.

If an equity firm came to me with three investments with expected returns of 7% each and i have £150,000 to invest I am going to (if I believe their numbers to be accurate) invest equally £50,000 in each to best spread my exposure. The equal EV makes it nonsensical to have more exposure in one over the other.

You done the opposite, you calculated basically equal EV in Ratajski and Joyce but staked 2.5x on Ratajski compared to Joyce.

[vizIIsto]

"If you like to bet with high risk, high reward, a.k.a. the SuperSwede system"

That is the opposite of what I do and suggested, I am the most cautious bordering on boring "gambler" in history but made a living off realising EV in another field for almost 15 years.

I was't suggesting you raised all your stakes to the Rataski level, I was suggesting your Ratajski level was too high and you were over exposing yourself.

If an equity firm came to me with three investments with expected returns of 7% each and i have £150,000 to invest I am going to (if I believe their numbers to be accurate) invest equally £50,000 in each to best spread my exposure. The equal EV makes it nonsensical to have more exposure in one over the other.

You done the opposite, you calculated basically equal EV in Ratajski and Joyce but staked 2.5x on Ratajski compared to Joyce.

I think this is a discussion on the topic of risk management more than on the subject of the viability of trying to maximize profit. But I digress, you have made your point. However I am going to try out my system, since I am of the belief that maximizing the chance of profit is more important than maximizing the profit on itself.

[SuperSwede]

"If you like to bet with high risk, high reward, a.k.a. the SuperSwede system"

That is the opposite of what I do and suggested, I am the most cautious bordering on boring "gambler" in history but made a living off realising EV in another field for almost 15 years.

I was't suggesting you raised all your stakes to the Rataski level, I was suggesting your Ratajski level was too high and you were over exposing yourself.

If an equity firm came to me with three investments with expected returns of 7% each and i have £150,000 to invest I am going to (if I believe their numbers to be accurate) invest equally £50,000 in each to best spread my exposure. The equal EV makes it nonsensical to have more exposure in one over the other.

You done the opposite, you calculated basically equal EV in Ratajski and Joyce but staked 2.5x on Ratajski compared to Joyce.

I think this is a discussion on the topic of risk management more than on the subject of the viability of trying to maximize profit. But I digress, you have made your point. However I am going to try out my system, since I am of the belief that maximizing the chance of profit is more important than maximizing the profit on itself.

You are effectively saying you'd rather win less money overall in order to have less peaks and troughs which is a strategy that makes no sense long term (assuming it's a long term strategy as if profitable why stop)

Good luck though, even if I don't agree with all the logic it's at least good to see a calculated approach.

[vizIIsto]

You are effectively saying you'd rather win less money overall in order to have less peaks and troughs which is a strategy that makes no sense long term (assuming it's a long term strategy as if profitable why stop)

Good luck though, even if I don't agree with all the logic it's at least good to see a calculated approach.

Well, yeah, I am effectively trying to say that. Although I do have to say I would use this approach in the earlier stages of betting if I were to bet more money. When my bankroll has grown enough, I would take more risks since I have more room for bad luck.

Another question for you:
What/Who do you think is better, someone who takes a calculated approach but has an imperfect strategy, or someone who bets solely on their own knowledge and expertise? A.k.a. who wins more, a calculated better with inaccuracies like me or a darts pundit better who understands the concept of betting better?

[SuperSwede]

You are effectively saying you'd rather win less money overall in order to have less peaks and troughs which is a strategy that makes no sense long term (assuming it's a long term strategy as if profitable why stop)

Good luck though, even if I don't agree with all the logic it's at least good to see a calculated approach.

Well, yeah, I am effectively trying to say that. Although I do have to say I would use this approach in the earlier stages of betting if I were to bet more money. When my bankroll has grown enough, I would take more risks since I have more room for bad luck.

Another question for you:
What/Who do you think is better, someone who takes a calculated approach but has an imperfect strategy, or someone who bets solely on their own knowledge and expertise? A.k.a. who wins more, a calculated better with inaccuracies like me or a darts pundit better who understands the concept of betting better?

I like to think I am a hybrid of the two but given the choice of the options you gave i'd side with your approach simply because the other will be more likely to drastically increase stakes and chase losses.