JJ's Blog

So this is my second year playing fantasy basketball. I find myself falling into the same routine that I went through with poker. That is, phase 1, play the game badly, but have fun. Phase 2. learn the game and try to play better.

So during the first fantasy B-Ball season, my only strategy was to try to assemble a team that would dominate a few of the scoring categories and might luck out and win another couple of random categories. After all, you just need to win 5 out of 9 categories, and you win your head-up match-up for the week. So last year, I tried to put together a team of 3 point shooting big men. That way, I’d win the following categories: 3PTS, Rebounds, and FT%. I might get lucky and win PTS and blocks.

It didn't work so well. Not sure if it was a bad strategy or the fact that some of my big men turned into real duds (i.e., Andre Kirilenko).

This year I read up a little before the draft and tried to do a better job drafting a “balanced team”.

Fantasy basketball is somewhat similar to poker. If possible, you should try to start with a good hand (i.e., draft good). Then, you need to play the hand well (make a few trades or drop/adds to keep your scoring balanced). Then, in the end, you need to get a little lucky (i.e., pick up that dark-horse player –i.e., Deron Williams, that exceeds everyone’s expectations).

So my drafting strategy was as follows:
1. Draft a balanced team by position. This is key. You really can’t afford to have any open roster spots throughout the week because you don’t have certain positions available. You need to have a balance of Point Guards, Shoot Guards, Small Forwards, Power Forwards, and Centers. One way to do this is to give some preference to players that can play multiple (3) positions. But at a minimum, you need to make sure you have each of the positions covered and balanced.
2. Draft a team that provides balanced scoring across all of the statistical categories.
3. Draft a team of players that….plays. The player must not be injury prone (i.e, played 80 games last year) and they must log big minutes.

So here’s what I did.
Step 1. I built a spreadsheet that would sum up all of my teams statistics as I added each player during the draft. As I selected a player, I would copy/paste his stats from last year into the spreadsheet and watch the totals line change.
Step 2. Compare my team’s totals statistics to statistical winners from last year’s yahoo fantasy basketball league. The spreadsheet would automatically calculate a % deviation statistic for each category. Then, I graphed the % deviation. This way, I had a constantly updated graph of how my team was “shaping up” in each of the categories as I added each player – compared to last year’s category leaders. If I was doing well in the 3pt category (the bar was high), but doing poorly in the steals category (the bar was negative), my next draft selection would be used to address steals.
Step 3. I kept track of each of the positions my players could play. If I drafted a player that could play SF, F, and SG, I would count this player in all three positions. Then, I kept a running sum of players by position. That way, I could easily keep balance across the positions.

So, in the first two rounds, I just selected the best available players. Then, in the last 11 rounds, I would add players that addressed the needs that were indicated on the spreadsheet graph and that would keep balance on my roster. Before I selected a player, I’d just make sure they were not injured, played a “healthy” number of games last year, played a “healthy” number of minutes last year, and could play a variety of positions.

Of course the flaw with this strategy is that is based completely on last year’s performance (both for the players and fantasy teams). This strategy does not take into account situations like a player’s declining health, rookies, a player’s potential, or special situations (i.e., Kobe Bryant’s tenuous situation with the Lakers).

I think I drafted a pretty good team. I guess we’ll see if it can outscore my opponents.

So I just started playing online poker.

In reality, I've played online poker for the past couple of years, but it was usually for a few hours on the weekend, here and there, nothing too serious. I think I deposited a total of about $200 online at different sites. I'd play a variety of games, cash games, tourneys, etc... My main objective was to learn all of the different games. If I was hosting a Stud game at my house, I'd play nothing but Stud for 2 months. Then, if I was hosting an Omaha game at my house, I'd play PLO for 2 straight months. This is a good way to get some "volume" in on some of the mixed games so you don't completely embarrass yourself. I think I played at max stakes of $2/$4 limit H.O.R.S.E. Eventually, it was all gone.

So when I say I'm starting to play online, I mean that I'm starting to play with profit as the only objective. I'm approaching this kind of like a part time job (around 15 hours a week) and a small business (i.e., I need to invest a lot of time and effort initially and only expect a profit several months down the road).

It's a pretty low risk experiment, because if I'm no good, I'll only loose a little money. If I can actually beat the games, then I'll make a little money - down the road. The biggest downside is that I may waste 200 hours or so before I figure out that this is a bad idea.

Anyway, I'm playing NLHE. My question is what is the minimum payroll requirements needed?

Here's what I'm thinking. I'll play with 200 big blinds (double stack) and I'll move up when I have 50 stacks and I'll move down when I have less than 30 stacks.

First, I'm absolutely giddy about the WSOP televising 6 hours of H.O.R.S.E. My tuesday nights are locked up for the next 3 weeks.

Second, order is about to be restored in Huskerland. The legend, the original T.O., Tom Osborne is been tapped to bring order to the House of Husker. http://sports.yahoo.com/ncaaf/news;_ylt=AoUt5a1imHVl72.kN5BkKXs5nYcB?slug=ap-nebraska-osbornereturns&prov=ap&type=lgns

Third, is there a funny show on network TV than The Office? I think not.

You know what playing about 10,000 hands of NLHE online has taught me….

Volume matters
When you can cram a whole bunch of hands into a small window of time, it is much more likely that you’ll experience a high number bad suckouts by your opponents. You are also able to put a bunch of bad beats on your opponents, and you are able to experience a really high number of hands that play out just the way they are “supposed to” – i.e., the best hand holds up after all of the money goes in.

When you play a live game session, you might get dealt AA 3 times. If you get no action 2 times and then get stacked when your opponent hits a miracle flop, if you’re like me, you assume that you are on some horrible run and the other players are really bad. If, by magic somehow, you were able to compress about 10 of these sessions into the same time window (as you can in online poker), you get a much more balanced view of the distribution of outcomes with AA (i.e., you get dealt AA 30 times, you double up 4 times, it’s folded around to you 10 times, you win with a continuation bet 6 times, you win with a turn bet 4, you have to fold to a dangerous board 3 times, and you get stacked 3 times).

So I guess I’m beginning to appreciate the variation and swings of NLHE a little more – and maybe I’m becoming a little more tolerant.

Profit Centers
I’ve also started to learn where profit comes from in NLHE for me. In business, you can make a profit by selling lots of items with a very small profit margin (i.e., newspaper) or you can sell a few items with a huge margin (i.e., yachts). In poker, you get an opportunity for both. Poker Tracker will tell you some of this info. But I’ve learned that I win $$ primarily from the following 4 profit centers. These are listed in descending order of profit margin.

1 – I flop a monster hand that is well disguised and I stack someone holding premium starting cards. A hand came up last night that is a good example. I held 4-6 suited in the big blind and I was able to stack someone holding AA who min raised from the button. After the flop comes 5-7-A, it goes check, check, and a 3 comes on the turn, and we re-raise each other until we both have a double stack in the middle of the table – Gigantic margin
2 – I flop a set – Big Margin
3 – I start with a premium hand and it holds up – Medium Margin
4 – I play a mediocre hand in late position and catch a flop or I win with a bluff – small margin (4 or 5 Blinds)

When I multiply the margin from these 4 situations by the frequency of them happening, I discovered that I’m making most of my money from profit center #4 – that is, playing aggressively from late position. The following hand illustrates this. I’m in the cutoff with J-9o. Action is folded to me. I raise it 4x the BB. The SB and BB defend. The flop comes A-10-8. SB and BB check. I bet 2/3rds of the pot and win it. This is a pretty low risk, standard play. The margin is pretty small (8 big blinds), but it happens so frequently that it has become my #1 profit center.

You can tell I'm missing poker.

So an interesting hand came up the other night playing euchre. I’m not a euchre player. I have absolutely zero skill at this game. But I thought we might be able to use some of the same concepts that we use in poker to analyze the correct play.

Quickie overview of Euchre
So we are playing 4 handed (which I guess is standard – even though there are 2 handed or 6 handed variations). Euchre is played with partners. The goal is to score more points than your opponent’s team (a race to 10 points is a game). You score points by taking “tricks” as in other suit game like hearts or spades. The game uses 24 cards (A through 9). The rank of the cards is the same as in poker except the Jack (aka right bower) is higher than an Ace for the trump suit. Another wrinkle is that the Jack that is the same color but different suit as the trump is the second highest card. This is the left bower. For example, when trump is hearts the ranks are Jh, Jd, Ah, Kh, Qh, 10h, 9h. The ranks for the other suits (non-trump) are standard poker ranks. For example when trump is hearts, the ranks for spades and clubs would be A,K,Q,J,10,9. The rank for diamonds would be A,K,Q,10,9 (notice that the J of diamonds is not used since it would be a left bower for trump).

You are dealt 5 cards each. The remaining 4 cards are put into a discard pile. Then, the top card from the discard pile is turned face up. Each player, in order, has the option of declaring that card’s suit as trump. If that card’s suit is declared as trump, the dealer puts that card into their hand and discards another card. If everyone passes (checks) on declaring trump, each player in turn has the option to declare any other suit as trump. If no one takes their “option” to declare trump, the dealer, who is last to act, must declare trump (this is the variation called “stick the dealer").

You can score points in the following ways:
* When you order trump and take 3 tricks or 4 tricks – 1 pt
* When you order trump and take 5 tricks – 2 pts
* When your opponents order trump and fail to take 3 tricks – 2 pts (called “getting euchred”)
* When you order trump and “go alone” (do not use a partner) and take 3 tricks or 4 tricks – 1 pt
* When you order trump and “go alone” and take 5 tricks – 4 pts
* When you order trump and “go alone” and fail to take 3 tricks – opponents score 2

The Setup
The opponent to my right (villain 1) is dealing and after it is passed around, she orders up hearts as trump when she picks the 10h off the pile. She then proceeds to take all 5 tricks when she plays 5 successive trump cards – no one else had trump. As it turns out, they whomped us 10 to 0 in that game.

Like poker, it is somewhat expected and traditional to berate a player for making a stupid play after they beat you. So me and my partner started give our opponent some shit for not going alone and trying to score 4 points. She defended by saying that since she had neither bower, the correct play is to use a partner.

Let’s take a look at the situation.

Since Villain 1 had 5 trump cards, there are only 2 trump remaining. These 2 remaining trump (the 2 bowers – highest ranking cards) are somewhere in my hand, my partners hand, my other opponent’s hand, or in the 3 cards left in the discard pile.

Like in poker, I think we can just construct a decision matrix with scenarios and probabilities. The way I see it, there are about 7 possible scenarios. I’ve summarized them below

Scenario 1: If both bowers (Jacks) were held by me or if both bowers were held by my partner, we would take 2 of the 5 tricks – since one person on my team would have the 2 highest ranked trump cards. In this situation, villain would score 1 point by ordering trump and going alone or by ordering trump and using a partner. There is no difference in the outcome. Going alone or taking a partner would net 1 point.

Scenario 2: If both I and my partner held a bower each, there could be 2 “2nd order” outcomes. One, my partner could lead with a non-trump card which would be trumped by my opponent. She would then lead with trump and would “draw” both bowers from me and my partner on the next card. That’s one trick for us. Then, our opponent would take the rest of the tricks (4 in total). Two, if my partner leads with his bower, it would draw my bower, and once again, we would take 1 of the 5 tricks. In either case, my opponents would take 4 tricks. So, going alone or using a partner in this case would result in our opponents scoring 1 point.

Scenario 3: If both of the bowers were held by Villain 2 (Villain1’s partner), they would take all 5 tricks. In this case, they would score 2 points for taking all 5 tricks and using a partner, but they would have scored 4 points by going alone. In this case the better play is to go alone.

Scenario 4: If both bowers were buried in the discard pile, villains would take all 5 tricks. In this case, they would have scored more points by going alone (4) vs. using a partner (2).

Scenario 5: Either me or my partner hold a bower and the other bower is buried. In this case, my team would take 1 trick. Our opponents would take 4 tricks. They would score 1 point either way (going alone or taking a partner)

Scenario 6: Either me or my partner hold the right bower and our opponent holds the left bower. In this case, my team would take 1 trick. Our opponents would take 4 tricks. They would score 1 point either way (going alone or taking a partner)

Scenario 7: Either me or my partner hold the left bower but our opponent holds the right bower. In this case my team would take 0 tricks. Our opponents would take 4 tricks by going alone and would score 1 point, but they would take all 5 tricks and score 2 points by using a partner.

Summary
Out of all 7 of the possible scenarios, there are 4 (#1, #2, #5, and #6) that there is no difference in the points that result from the choice to go alone or to take a partner. There are 2 scenarios (#3, #4) where going alone results in 2 more points than taking a partner. There is one scenario (#7) where taking a partner results in one more point than going alone. Let’s assume for now that each scenario has the same probability of occurrence (1 out of 7).
------------------------------------------------------------------------------------
Scenario, Probability, Alone Pts, w/Partner Pts, extra Pts
------------------------------------------------------------------------------------
.....1...............1/7...............1..............1................0
.....2...............1/7...............1..............1................0
.....3...............1/7...............4..............2................2
.....4...............1/7...............4..............2................2
.....5...............1/7...............1..............1................0
.....6...............1/7...............1..............1................0
.....7...............1/7...............1..............2...............-1
==========================================

Then, the equation for points (EV) for going alone would be.

4 out of 7 chances you score 0 extra points + 2 out of 7 chances you score 2 extra points + 1 out of 7 chances you score 1 less point
Or
(4/7 * 0) + (2/7 * 2) + (1/7 * -1) =

.43 pts – this would suggest the best play is to go alone when you hold 5 trump with no bowers.

In reality, the probabilities of each of these situations are not identical. I think we can simplify the problem by just determining the probability for scenario 3, 4 and 7.

For scenario 3, the probability that our villain (knowing the 5 cards that are in her hand) can assign to the likelihood that her partner holds both bowers is 5 out of 19 unseen cards multiplied by 4 out of 18 unseen cards. This is similar to the way we calculate the probability of catching runner-runner flush cards in holdem. In this case it is 6%.

For scenario 4, the probability that both bowers are in the discard pile would be 3/19 x 2/18 or 2%

For scenario 7, the probability that either me or my partner hold the left bower and our opponent holds the right bower is 10 chances out of 19 unseen cards x 5 chances out of 18 cards = 15%

So the revised probability matrix might look like this

------------------------------------------------------------------------------------
Scenario, Probability, Alone Pts, w/Partner Pts, extra Pts
------------------------------------------------------------------------------------
..1,2,5,6........77%.............1..............1................0
.....3...............6%...............4..............2................2
.....4...............2%...............4..............2................2
.....7...............15%.............1..............2................-1
==========================================

So the revised EV is
(77% * 0) + (6% * 2) + (2% * 2) + (15% * -1) = .01 extra pts - this would seem to suggest that there is no real difference in going alone or taking a partner when you hold 5 trump cards with no bowers in Euchre.

As always, let me know if I got this stuff right or wrong.