Y. Takahashi vs A. Chang — prediction

Takahashi carries a real rating advantage — 1568 Elo against Chang's 1495 — which the model converts into a 60% win probability. This is a modest gap in a soft ITF market, not a dominant one, and it reflects overall level rather than any specific surface or style data, since neither player has serve, return, or surface percentages on record here.

The rest picture cuts against the favorite. Takahashi is playing on just 1 day of rest and has logged 3 matches in the last 14 days, a workload that can blunt legs and focus in a decisive third set. Chang, by contrast, has had 4 days to recover, though he is also carrying the residue of a deep run — he reached the semifinals at this same M15 Bali event only 4 days ago, so his freshness edge is not absolute.

Taken together, these two fatigue signals partially offset each other: Takahashi's tighter turnaround is a clearer, more immediate drag than Chang's slightly older deep-run fatigue.

Neither player is in strong form, but Chang's last 10 results (5 wins, including a 3-match winning streak mid-stretch) look marginally better than Takahashi's (4 wins, more scattered losses). This is a shallow signal — win/loss strings don't capture opponent quality — but it adds a small tilt toward Chang rather than reinforcing the favorite's Elo edge.

The market prices Takahashi at an implied 69% (odds of 1.44), well above the model's 60% estimate, producing a -13.3% expected value on the favorite. That gap says the market is more confident in Takahashi than the data-driven estimate supports — this is not a case where being the favorite equals value.

It's also worth flagging that this projection comes from a soft Elo-based method built for lightly-tracked Challenger/ITF matches, so its edge over the market is unproven. Combined with the negative EV, there is no statistical case for backing the favorite here at these odds; if anything, the numbers lean toward caution rather than conviction on either side.