From the Betting Ring to Winner’s Enclosure: Following the Money with SmartForm

With the SmartForm database at our fingertips, we can test some hypotheses on the profitability of following the money, in particular any clues that can be picked up from horses that have been backed on their previous outings.

In this case, we will take as a given that significant contraction of odds indicates a better chance of winning, not least because the implied probability of winning becomes higher – instead, our mission here is to look at what is required to analyse any potential path to profitability from following previous or current market movements, a very different enquiry from simply looking at winner strike rates.

Decoding the Betting Text: The Genesis of Our Inquiry

Every seasoned punter knows that the betting market is a dynamic entity, reflecting collective wisdom, insider insights, and sometimes, mere speculation.

The betting_text field in the SmartForm database captures the ebb and flow in bookmaker starting price, which approximately chronicles the shifts in odds offered from the on course opening show (roughly from when horses appear in the parade ring) right up to the off and the official starting price, appearing just before the off, and offering a useful proxy for late betting sentiment.

Using string matching in SQL on the Smartform database, we parsed this text field to extract structured indicators for:

op_string – the opening price
sp_string – the starting price
tchd_string – any better price touched in between

These values were standardised (e.g. “Evens” converted to “1/1”) and transformed into implied probabilities, allowing us to calculate percentage price moves from open to off — both for last time out (LTO) and for the current race.

The resulting database table can be generated in a minute or so for hundreds of thousands of runners and then combined with core Smartform features in other tables — including changes in official rating and class — to explore whether backing activity in a previous run could offer a profitable angle when viewed in the context of other race factors.

Establishing the Baseline: Market Movements and Their Implications

Before diving deeper, it's essential to understand the implications of market movements alone in terms of profitability.

Figure 1: ROI by Current Price Movement and SP Probability Band

The heatmap above illustrates the Return on Investment (ROI) across various bands of Starting Price (SP) probabilities (on the X axis) and the magnitude of price movements from the off (Y axis).

Notably, runners with significant price shortening often yield higher ROIs, however by far and away the greatest predictor is the implied probability itself.

Neither takes us into profitability to prices returned at SP alone (a high bar for any ROI calculation) but the closest we can get (next to neutral colour) is a horse that has been significantly backed and returns at odds on (> 50% implied probability band).

Figure 2: Win Rate vs Implied: Current Market Movement Signal

This companion heatmap shifts focus from profitability to predictiveness — showing how actual win rates compared to the market's implied probabilities. While the overall structure mirrors Figure 1, the key insight is that significant shortening correlates with a better alignment between expectation and outcome.

However, even the most heavily backed runners rarely exceed the win rate implied by their odds, underlining the efficiency — but also the tight margins — of the SP market.

Now it’s time to test some hypotheses using the new table, namely whether or not we can profit from following money laid on the horse’s last time out run, in the hope that this was indicative of future promise.

The next heatmap brings together two key components: how strongly a horse was backed last time out (shown on the Y-axis) and its implied probability today based on SP (on the X-axis).

All runners included had a drop in both official rating since their last race and were racing at the same or lower class, as part of our core hypothesis test.

Figure 3: ROI by Strength of LTO Backing and SP Probability

The picture here is sobering — much of the ROI landscape remains a shade of red, reaffirming the challenge of beating SP returns over time.

However, there’s a clear trend: the more a horse was backed LTO, the less red the ROI becomes.

In particular, the lightest tones (i.e. best returns) occur when LTO price movement exceeded 20%.

We might conclude that market support in a horse’s previous outing does carry some signal — but on its own, even when combined with favourable rating/class drops, it doesn’t yet offer a standalone path to profitability.

Still, it points us in a useful direction: LTO market strength appears to be a meaningful filter and could be even more potent when cross-referenced with other form indicators.

Let’s continue this strand by narrowing in on one race type – Flat Turf racing – and additionally considering an indicator of last time out performance alongside our other criteria – in this case simply whether the horse was placed last time out.

Figure 4: Flat Races: ROI by Strength of LTO Backing and LTO Placement

In this heatmap, we apply tighter filters to test whether prior market confidence — i.e., money coming for a horse last time out (LTO) — becomes a more useful signal when coupled with a solid performance on the track. Specifically, we now restrict the sample to:

Flat handicap races
Horses whose official rating dropped from last time out to the current race
Same or easier class than LTO
Backed in the betting LTO
Then split by whether or not the horse placed last time out (LTO)

The heatmap is again broken down by the strength of the LTO price movement (rows) and the SP probability for the current race (columns).

ROI is shown in colour.

Now we see a clear distinction between ROI from horses on the Left panel (Not Placed LTO) where there is little to suggest following money from last time out alone yields any edge if the horse didn’t perform on the day to horses on the Right panel (Placed LTO)
where it gets interesting.

A clear cluster of pale shades to green begins to emerge in the 10–30% SP probability bands for horses that were:

Strongly backed last time out (SPs 20% to 50% + lower than their opening show)
Dropped in rating and remained in the same class or lower as last time
Showed some form by placing LTO

The best ROI (green on the heatmap) occurs for those that are mid-range horses in the market today, not favourites, which is perhaps a key consideration when looking at ROI alone – the shorter the price the harder it will be to eek positive returns over a larger a sample, but outsiders are also outsiders for a reason. In this sample, there is an ROI sweet spot in the 3/1 to 20/1 range.

These findings set the stage for our final layer: bringing in today’s market moves alongside these signals to see if that sharpens the angle even further.

Figure 5: ROI by LTO Placement × Backing Strength × Current SP Movement

This final heatmap brings together the full story so far — combining three signals:

Support in the market LTO (price movement last time out)
A solid LTO run (Placed vs Not Placed)
Current price activity (Shortened, Same, or Lengthened)

We now facet the heatmap by both LTO placement status (rows) and today’s SP movement (columns), to assess whether market behaviour today reinforces or weakens the LTO signals we've uncovered.

Interpretation and conclusions

The most encouraging signs emerge in the bottom-right quadrant — horses that were:

Backed last time out
Ran into a place
And are being backed again today

In this subset, we see multiple pockets of ROI turning green — especially in the 20–50% SP probability bands, aligning with our earlier observation that value is often found among mid-market runners, albeit this time returns are shorter, as we are by definition focusing on horses that have already contracted in the price that we are using for ROI calculations.

Even in the “Same” or “Lengthened” price movement columns, there are isolated greens — suggesting that not all today's market drift is damning, particularly if the LTO profile is strong.

By contrast, horses that were backed LTO but failed to place offer no consistent signs of value, regardless of today’s move — echoing our earlier conclusion that some indicator of performance to match the money matters.

There are far too many permutations to cover to give a comprehensive overview of following the money across all race types and conditions, but this brief overview shows the utility of analysing significant market moves at the business end of the days’ betting.

We even managed to identify pockets of value at starting price – which bodes very well for potential ROI at Betfair or Tote prices.

It’s a reminder that the market isn’t just noise — but also not infallible. And where patterns repeat, edges can emerge.

Colin Magee