Forecast evidence, with current live performance first.

Gramm previously published historical backtests that suggested 1-3% MAPE. Current live production evidence does not reproduce those numbers, so this page separates live scorecards, methodology, and research archive.

7.61%
Gramm production h24 MAPE, measured ISO-reference grids
2.70%
ISO h24 reference average on the same measured grids
Hold
Promotion status until a candidate beats live references

Current production h24 MAPE.

Last updated May 22, 2026. This is the public production scorecard. It intentionally shows the current gap to ISO rather than hiding weak performance behind research history.

Grid
ISO h24
Gramm h24
Decision
CaliforniaCAISO
3.52%
8.13%
4.61 pp behind ISO
TexasERCOT
2.50%
8.24%
5.74 pp behind ISO
Mid-Atlantic & MidwestPJM
n/a
7.65%
reference pending
Central U.S.MISO
2.46%
6.11%
3.65 pp behind ISO
New YorkNYISO
1.80%
7.76%
5.96 pp behind ISO
New EnglandISO-NE
n/a
14.49%
reference pending
Great PlainsSPP
3.20%
7.84%
4.64 pp behind ISO
ISO-NE and PJM ISO h24 references are marked pending in this table until the canonical paired-reference record is complete.

Standard Metric rules.

Paired rows only

Forecasts are compared to actuals at the same grid, horizon, issue-time contract, and target timestamp. Unpaired rows do not enter the score.

ISO reference visible

When the ISO h24 reference is available, it is shown beside Gramm production. Missing ISO references are marked pending rather than filled by proxy.

Shadow before promotion

A candidate can look promising in offline tests and still fail live. Promotion requires shadow evidence, not just a historical validation score.

Historical research archive, not production claim.

These values explain why the research direction looked promising. They should not be used in sales material as current forecast quality until reproduced under live Standard Metric evaluation.

Grid
Archive ISO
Archive model
Archive reduction
CaliforniaCAISO
4.80%
2.82%
41.3%
TexasERCOT
5.10%
2.45%
52.0%
Mid-Atlantic & MidwestPJM
3.90%
1.75%
55.1%
Central U.S.MISO
4.30%
1.74%
59.5%
New YorkNYISO
4.10%
2.21%
46.1%
New EnglandISO-NE
3.70%
3.18%
14.1%
Great PlainsSPP
4.60%
3.30%
28.3%

Evaluation path

Use the same metric for your own audit.

Annual, quarterly, and pilot audits should use paired h24 MAPE rows, with grid, period, horizon, and reference source declared before scoring.

Request a benchmarkGet API key
  1. 01

    Review live scorecard

    Start with the current h24 table, including gaps to ISO references and pending reference rows.

    Open scorecard ->
  2. 02

    Request scoped benchmark

    Declare grid, horizon, audit period, and decision context before any comparison is scored.

    Request benchmark ->
  3. 03

    Test API payloads

    Create a free key or inspect the docs, then test responses against your integration shape.

    View API docs ->
  4. 04

    Move to procurement or paid plan

    Use the trust, SLA, pricing, and enterprise pages when the benchmark is ready for review.

    Open procurement ->