A very special Happy Mothers Day goes out to all of the women that birthed our Armchair Managers. Why am I writing this extra post instead of spending time with my mother on this day? Well as we have seen in the past, the more that I probably shouldn't be focusing on fantasy baseball, the more I do. [Spain, finals week, etc ...] So today I have decided to give a small series of articles on my methodology. I know I have done some shorthand stuff in the past, but due to Dave's suggestion, I will use this as my definitive explanation.
This article will deal with perhaps the heart of all my statistics: Predicted Winning Percentage (PredW%). Its name attempts to do exactly what it should; give a better indicator of what your W% should be, based on how you have done compared with the rest of the league, opposed to just your head to head opponents. So earlier in the season, your Actual Win Percentage (ActW%) and PredW% may be a bit different (indicating bad/good luck with who you have been matched up against) but as time goes on, these numbers have shown to converge toward each other.
By comparing your weekly totals with the rest of the league rather than only one team, we get a better indicator of how well your team is within the league, simply enough. So let's go through the math of it, shall we? I first look at everyone's category stats for a single week. Let's observe Week 4:
| Week | Raw | R | HR | RBI | SB | AVG | OPS | W | L | SV | K | ERA | WHIP |
| 4 | Make It Rain (TFP) | 37 | 13 | 34 | 6 | 0.29 | 0.887 | 5 | 4 | 4 | 57 | 3.52 | 1.14 |
| 4 | Enter Sandman | 24 | 5 | 27 | 3 | 0.231 | 0.648 | 5 | 5 | 4 | 76 | 3.92 | 1.38 |
| 4 | Sexual Napalm | 32 | 8 | 30 | 5 | 0.306 | 0.888 | 5 | 2 | 0 | 44 | 2.78 | 1.25 |
| 4 | Maewyn Succat | 32 | 6 | 37 | 7 | 0.357 | 0.99 | 10 | 4 | 4 | 93 | 2.83 | 1.21 |
| 4 | La Furia Roja | 35 | 7 | 35 | 4 | 0.265 | 0.774 | 1 | 6 | 5 | 64 | 5.83 | 1.55 |
| 4 | The Master Baker | 33 | 14 | 41 | 3 | 0.278 | 0.909 | 5 | 0 | 4 | 60 | 1.45 | 1.11 |
| 4 | Drunk Hot Girl Scout | 34 | 4 | 16 | 3 | 0.245 | 0.701 | 4 | 2 | 0 | 46 | 6.15 | 1.58 |
| 4 | Winston Wolfe | 30 | 8 | 26 | 3 | 0.278 | 0.853 | 3 | 2 | 2 | 54 | 3.3 | 1.31 |
| 4 | Dmitri Petrovich | 30 | 8 | 33 | 2 | 0.229 | 0.694 | 8 | 4 | 4 | 94 | 2.76 | 1.12 |
| 4 | Joe's Torn Ligament | 32 | 3 | 22 | 10 | 0.251 | 0.691 | 8 | 4 | 6 | 78 | 4.11 | 1.38 |
Next, I don't pay any attention to who plays whom for the week. I simply look at each category for each team and ask, "How many other teams would I beat this week for this category?" For example, look at the HR category for Make It Rain (TFP). He had 13 HRs on the week. If we were to look at how many other teams he would have beaten, we can see that since he had the 2nd most HR, he would have beaten 8 out of the other 9 teams, good for a winning percentage of .889 for that category. However, as luck would have it, Make It Rain (TFP) played against Joe's Torn Ligament this week (which was randomly determined by the scheduling). So even though he had the 2nd best HR count in the league, he got 0 points. My system gets rid of the matchup situation and awards people wins compared to the league. So I do this for every team in the league and come up with individual W% for each category for each team:
| WeekRank | R | HR | RBI | SB | AVG | OPS | W | L | SV | K | ERA | WHIP | PredW% |
| Make It Rain (TFP) | 1.000 | 0.889 | 0.667 | 0.778 | 0.778 | 0.667 | 0.500 | 0.389 | 0.556 | 0.333 | 0.444 | 0.778 | 0.648 |
| Enter Sandman | 0.000 | 0.222 | 0.333 | 0.278 | 0.111 | 0.000 | 0.500 | 0.111 | 0.556 | 0.667 | 0.333 | 0.278 | 0.282 |
| Sexual Napalm | 0.444 | 0.667 | 0.444 | 0.667 | 0.889 | 0.778 | 0.500 | 0.778 | 0.056 | 0.000 | 0.778 | 0.556 | 0.546 |
| Maewyn Succat | 0.444 | 0.333 | 0.889 | 0.889 | 1.000 | 1.000 | 1.000 | 0.389 | 0.556 | 0.889 | 0.667 | 0.667 | 0.727 |
| La Furia Roja | 0.889 | 0.444 | 0.778 | 0.556 | 0.444 | 0.444 | 0.000 | 0.000 | 0.889 | 0.556 | 0.111 | 0.111 | 0.435 |
| The Master Baker | 0.667 | 1.000 | 1.000 | 0.278 | 0.611 | 0.889 | 0.500 | 1.000 | 0.556 | 0.444 | 1.000 | 1.000 | 0.745 |
| Drunk Hot Girl Scout | 0.778 | 0.111 | 0.000 | 0.278 | 0.222 | 0.333 | 0.222 | 0.778 | 0.056 | 0.111 | 0.000 | 0.000 | 0.241 |
| Winston Wolfe | 0.167 | 0.667 | 0.222 | 0.278 | 0.611 | 0.556 | 0.111 | 0.778 | 0.222 | 0.222 | 0.556 | 0.444 | 0.403 |
| Dmitri Petrovich | 0.167 | 0.667 | 0.556 | 0.000 | 0.000 | 0.222 | 0.833 | 0.389 | 0.556 | 1.000 | 0.889 | 0.889 | 0.514 |
| Joe's Torn Ligament | 0.444 | 0.000 | 0.111 | 1.000 | 0.333 | 0.111 | 0.833 | 0.389 | 1.000 | 0.778 | 0.222 | 0.278 | 0.458 |
So PredW% for each week, is just the average of each category. It just gives a number that shows, compared to the rest of the league for that week, how good you typically were for each category. So once we know how good a team is for a particular week, I simply average those out with every other week to give the current, all encompassing PredW% you all see in my WIRs. For instance, Make It Rain (TFP) had a PredW% for RBIs of .778 in Week 1, 1.000 in Week 2, .889 in Week 3, and .667 for Week 4 (as we can see above). So if we average them out, we should see that they should have won about .833 of their RBI categories. We can look at our Head to Head W% Stats and compare our predicted W% vs actual W% for this category. We see Make It Rain (TFP) is 2-2-0 in the RBI category on the year, or .500. Based on the first four weeks of the year, we can see they are a very good RBI team thus far, yet his actual W% is only .500. Therefore, I would expect these two numbers to become closer and closer as time goes on and luck thins out. He has been very unlucky thusfar in my opinion, and if he continues to put up similar RBI totals compared to the league, there's no doubt he will start beating teams in the RBI category more regularly and his actual RBI W% will climb.
So when I post PredW% each week in my WIR, it's basically an average on where you rank in each category each week compared to the rest of the league. I do have data from last year that shows that as the season goes on, this is a better predictor of future W% than actual W% at any given point throughout the year. So if we're looking for how a team will do in the future, look to my stats and not the league stats as a better indicator.
Anyway, if there's anymore that needs to be explained, call, reply, text, e-mail, wall post, or whatever you care to do to contact me and I'd be happy to settle any confusion.
Until next time people, pleasant fantasy baseballing.