Tracking Your Strikeout Prop Results: Spreadsheet Setup and Metrics
Available in US
The Bet Log I Wish I Had Started Five Years Earlier
For my first three years of K-prop betting, I had no tracking system. I remembered the wins. I forgot the losses. At the end of each month, I checked my sportsbook balance and called the difference my “results.” That number told me almost nothing useful – not which types of bets were winning, not where my model was miscalibrated, not whether my edge was growing or shrinking. I was driving with a fogged-up windshield and calling it navigation.
The day I built my first tracking spreadsheet was the day my K-prop betting became serious. Not better overnight – serious. The spreadsheet forced me to confront uncomfortable truths about my betting patterns, and those truths eventually made me profitable in a way that gut-feel betting never could.
What Your Tracking Spreadsheet Needs
A K-prop tracking spreadsheet does not need to be complex. Mine started with ten columns and has grown to fourteen over four years, but the core ten are what matter. Here is what each one captures.
Date, pitcher name, and opposing team are the identification columns – they let you find and filter specific bets later. The prop line (e.g., over 6.5) and the odds you received (e.g., 1.91 decimal) capture the bet itself. Your model projection (e.g., 7.2 projected Ks) records what your analysis said before you placed the bet. The actual result (how many Ks the pitcher recorded) captures the outcome. Win/loss is the binary result. Units staked and units profit/loss track the financial impact.
Those ten columns give you the foundation. The four I added later bring analytical depth: the closing line (what the odds were at first pitch, after all market movement), the edge grade (A/B/C based on my scoring system), the umpire assignment, and the game-time temperature. The closing line is particularly valuable because it lets you measure closing line value, or CLV – the difference between the odds you received and the final market price.
Do not over-engineer the spreadsheet at the start. Begin with the core ten columns and add analytical columns as your betting matures. A sprawling spreadsheet that you abandon after two weeks is worse than a simple one you maintain all season.
The Metrics That Actually Tell You Something
Raw win rate is the number everyone checks first, and it is the least useful metric on its own. A 58% win rate at -110 is excellent. A 58% win rate at -150 is mediocre. Context matters – and the metrics below provide it.
ROI (return on investment) measures your profit as a percentage of total units risked. If you staked 200 units over the month and profited 16 units, your ROI is 8%. This is the single most important performance metric because it accounts for both win rate and odds. A bettor who wins 52% at +120 average odds can outperform one who wins 60% at -150. ROI captures the full picture.
CLV (closing line value) measures whether you consistently got better odds than the market’s final price. If you placed a K over at 1.91 and the line closed at 1.83, you captured positive CLV – you got a better price than the market eventually settled on. Consistent positive CLV is the strongest indicator that your bets have genuine edge, because it means the market moved toward your position after you bet. Even if your short-term results are negative (variance happens), positive CLV strongly suggests long-term profitability. RotoWire’s documented K-prop approach – which posted a 76% win rate over a tracked season – was built on the principle of identifying mispriced lines before the market corrected them, which is exactly what CLV measures.
Edge grade distribution tells you whether your tiered staking is working. If your A-grade bets (highest conviction) are winning at 62% and your C-grade bets are winning at 51%, your grading system is calibrated well – you are correctly identifying which bets deserve more capital. If A-grade and C-grade bets win at similar rates, your grading criteria need refinement because they are not distinguishing between strong and weak edges.
Calibration measures whether your model projections match reality. If you project 7.0 Ks and the average actual outcome for those bets is 7.1, your model is well calibrated. If you consistently project 7.0 but the average outcome is 6.2, your model overestimates K totals and needs adjustment. Check calibration monthly rather than weekly – weekly samples are too small to distinguish signal from noise.
How Often to Review and What to Look For
I review my tracking spreadsheet on three cadences: daily, weekly, and monthly. Each review serves a different purpose.
The daily review takes thirty seconds. I log the previous day’s results – outcomes, units profit/loss, and any notes about unusual factors (rain delay, injury exit, unexpected lineup). The goal is data entry, not analysis. Do it immediately, before you forget the context.
The weekly review takes fifteen minutes. I check my rolling 7-day and 30-day ROI, look at my CLV trend (am I still getting good prices?), and scan for any patterns in my losses. Are my losing bets clustered around a specific type of matchup, a specific day of the week, or a specific odds range? These patterns are invisible in daily results but emerge over a week.
The monthly review is the deep dive. I calculate full-month ROI, CLV, win rate by edge grade, and calibration metrics. I compare my projection accuracy for different pitcher tiers (elite vs. average) and different opponent types (high-K vs. low-K lineups). I look for systematic biases: am I overestimating K totals for left-handers? Am I underestimating the impact of cold weather? Am I winning on A-grade bets but losing on C-grade speculative plays? These questions can only be answered with a month’s worth of data and a structured spreadsheet to query.
The monthly review is also where I decide whether to adjust my model. If calibration is off, I tweak the weightings (pitcher K-rate vs. opponent K-rate, recent form vs. season average). If one type of bet is consistently unprofitable – say, Sunday day-game overs – I either refine my approach to that category or eliminate it from my betting entirely. The spreadsheet is not a passive record. It is an active feedback tool that shapes my future decisions. For the strategic framework that generates the bets worth tracking, the K-prop strategy guide covers the full decision pipeline from game selection to bet placement.
What metrics matter most when reviewing K-prop performance?
ROI (profit as a percentage of units risked) is the most important overall metric. CLV (closing line value) is the strongest predictor of long-term edge – if you consistently get better odds than the closing price, your bets have genuine value. Calibration (how closely your projections match actual outcomes) tells you whether your model needs adjustment. Win rate alone is insufficient without context on the odds at which you are winning.
How many bets do I need before my K-prop ROI is meaningful?
A minimum of 200 bets is needed before ROI becomes statistically meaningful. Smaller samples are heavily influenced by variance – a 10-bet winning streak or losing streak can swing your ROI by 20 or more percentage points. At 200 bets, the variance smooths enough to reveal whether your edge is real. At 500 bets, the confidence increases further. Track from the first bet, but do not make model changes based on fewer than 100 results.
Written by the editors at mlb Strikeout Prop Bets.
