Figure 218 Oct 2026From “What is time series forecasting?”

A forecast comparison record

Write down the series, test period, horizon, benchmarks, measure and the form of the result before any method is scored.

Cite Embed

Figure 2A forecast comparison record

Source: Author's decision worksheet. The test period and horizon fields follow Makridakis et al. (2020), pp. 4 and 5 of the in-press version held; the benchmark field follows Makridakis et al. (2020), pp. 3 and 4 of the in-press version; the measure field follows Hyndman and Koehler (2006), pp. 682 to 686, and Makridakis et al. (2020), p. 5 of the in-press version. The series and result fields, and the rule in the footer, are this guide's. The fields are reader inputs; no company data is supplied.

PNG4:3, 1600 × 1200, with the title, the source line and the attribution set in the image isoglu-2026-what-is-time-series-forecasting-fig02.png

Source and evidence

Each line names a claim related to this exhibit, its source, and how it was checked.

B TSF-OWN-C3 A forecast comparison is written down as one record: the series with its unit and frequency, the test period, the horizon and its reason, the benchmarks (naive, seasonal naive and a simple average), the measure and its scaling, and how the result will be reported against each benchmark at each horizon, the form fixed before scoring and the result filled in after (G5 F6); illustrated with a hypothetical manufacturer's 40 spare-part lines with twelve months held back and one forecast origin per line, as in M4 (G5 N4), horizons of three and twelve months, the seasonal naive method, Naïve 2 and a plain exponential smoothing average as benchmarks, errors scaled per line in the MASE way, and a note that M4 left intermittent series out and that neither source tests methods built for them; untested.Author worksheet grounded in MSA20-C1, MSA20-C3, MSA20-C4, HK06-C2, HK06-C3 and HK06-C6

Grades: A, verified against the printed page of the primary source · B, primary source, text layer only · C, authoritative secondary · D, reported.

From the essay

Revenue operations & AI

What is time series forecasting?

A time series forecast extends a series from its own past. In the M4 competition, a hybrid and combinations led; pure methods trailed.

18 Oct 2026

Read the essay

This exhibit belongs to an essay: the numbers live here, the argument lives there.

Cite or embed

APA 7

Isoglu, S. (2026). A forecast comparison record [Figure]. isoglu.com. https://isoglu.com/insights/exhibits/isoglu-2026-what-is-time-series-forecasting-fig02/

BibTeX

@misc{isoglu-2026-what-is-time-series-forecasting-fig02,
  author = {Isoglu, Sinan},
  title = {{A forecast comparison record}},
  year = {2026},
  howpublished = {Figure 2, isoglu.com},
  url = {https://isoglu.com/insights/exhibits/isoglu-2026-what-is-time-series-forecasting-fig02/},
  note = {From the essay {What is time series forecasting?}}
}

Embed this exhibit

Copy this HTML to use the PNG exhibit on your site with its source line and a link back to the exhibit page.

<figure>
  <a href="https://isoglu.com/insights/exhibits/isoglu-2026-what-is-time-series-forecasting-fig02/">
    <img src="https://isoglu.com/social/isoglu-2026-what-is-time-series-forecasting-fig02.png" alt="Figure 2: A forecast comparison record" width="3200" height="2400">
  </a>
  <figcaption>Figure 2: A forecast comparison record. Source: Author&#39;s decision worksheet. The test period and horizon fields follow Makridakis et al. (2020), pp. 4 and 5 of the in-press version held; the benchmark field follows Makridakis et al. (2020), pp. 3 and 4 of the in-press version; the measure field follows Hyndman and Koehler (2006), pp. 682 to 686, and Makridakis et al. (2020), p. 5 of the in-press version. The series and result fields, and the rule in the footer, are this guide&#39;s. The fields are reader inputs; no company data is supplied. <a href="https://isoglu.com/insights/exhibits/isoglu-2026-what-is-time-series-forecasting-fig02/">View exhibit</a>.</figcaption>
</figure>