ARI@TB

This is a narrow, low-edge play: the model is recommending ARI +1.5 (-135) despite a model win probability of 36.5% versus a market implied probability of 36.8%, leaving a small value gap of -0.3%; the strength score sits at 52% which signals only modest conviction. The core angle is a runline cover — buying half a run — rather than finding hidden moneyline value, because the composite layers produce mixed signals and the market has drifted slightly toward the home side.

This market opened -190/160 and is currently -195/160; the home no-vig implied moved from 63.0% to 63.2% (toward home). That small drift — a two point move in no-vig implied probability — indicates slightly more money or respect for the home side since open.

The most realistic way this loses is straight up: Tampa Bay’s better overall record (46-33) and seasonal 58% win rate make a single-run cover unlikely if the Rays produce average offense.

No reported injury impact.

The projected ARI 3 - TB 4.5 final score fits the season scoring rates (TB avg 4.5 scored · 4.3 allowed; ARI avg 4.3 scored · 4.6 allowed) and maps to a one-run spread outcome.

The strength score of 52% reflects a narrow advantage driven by mixed layer contributions and a value gap of -0.3% between model and market.

I expect a low-to-mid scoring contest where Tampa Bay nudges out Arizona by a run, consistent with season scoring (TB avg 4.5 scored) and the model projection; final score ARI 3 - TB 4.5.

Bottom line: take ARI +1.5 (-135) as a small, situational play buying half a run; the edge is tiny and this is a play for ticket diversification rather than aggressive allocation. If you require a stronger value gap, this game should be passed on given the -0.3% value gap.

Shop the price: the game opened -190/160 and is now -195/160 — if you can get ARI +1.5 for -135 or better take it, but recognize the value gap is only -0.3%; consider limiting unit size or using the runline as a hedge to a small ML ticket (ML +160) and always shop for better juice/lines before committing.