PMCPS: How obligations are counted

Version 1.1, 2026-11-02

What this is. A statement of how the Institute counts obligations in the Prediction Market Consumer Protection Standards, published so that any figure the Institute states about PMCPS can be recomputed from the published text by someone who does not work here.

What this is not. This note is not part of any standard. It imposes no requirement on any person, operator or product, and it restates no requirement from PMCPS-C, PMCPS-M or PMCPS-T. Nothing in it can be complied with or breached. It covers the PMCPS family only; the skill-based charitable gaming family publishes its own note and the two are not interchangeable.


1. The two figures

Primary. Every occurrence of shall, must, shall not, must not or may not in the operative text, matching the longest form once. "shall not" is one occurrence, not a "shall" plus a "not".

Secondary, the strict-leaf figure. The number of paragraphs in the operative text that contain at least one of those terms and are not lead-ins. A lead-in is a paragraph that introduces the items beneath it, and the rule for when it counts is in section 3, which is the part of this note most likely to be got wrong.

Both are reported. The primary figure catches a drafter packing several duties into one paragraph. The secondary catches the reverse, one duty split across many. Publishing both makes the measure difficult to move in either direction without the other figure dissenting.

2. What is operative, declared by number

Exclusions are declared by section number, never by section title. A title rule looks equivalent and is not: it silently removes any operative section whose title happens to begin with an excluded word. The companion family found this in SBCGS-T, where sections 811 and 882 are operative, carry seven obligations between them, and are deleted by a rule excluding titles beginning "Application" or "Scope". PMCPS is not currently harmed by a title rule, and is declared by number anyway, because the protection should not depend on the accident of how a section was named.

PMCPS-C. Operative text is Part 100. Excluded sections: 101 Purpose, 102 Scope and relationship to the other Standards, 103 Definitions. Sections 104 through 107 are operative.

PMCPS-M. Operative text is Parts 200, 300, 400, 500, 600 and 700. Excluded sections: 201 Purpose, 202 Application, 203 Definitions. Every other section is operative.

PMCPS-T is reserved and not drafted, so it has no figure.

Also excluded, in both documents: everything outside the operative Parts, being the cover block, the prefatory note and the appendices; every advisory or explanatory note; and every line beginning Basis, which records sources rather than imposing anything.

Notes are excluded, and in the published HTML they are aside elements. This matters more than it sounds. The Institute drafts in markdown, where a note is a blockquote, and v1.0 of this note described the exclusion that way. The published pages a reader actually recomputes from contain no blockquotes at all: the renderer emits each note as <aside class="note">. A reader stripping blockquotes therefore strips nothing and counts the notes, arriving at a different figure from the one printed here.

How much it is worth: the drafting notes hold one obligation term at v1.2, in PMCPS-C section 104, and two at v1.3, the second being in the PMCPS-M section 301(d) note. So a reader who missed this rule would compute 206 and 132 at v1.2 against the 205 and 131 printed here, and 231 and 143 at v1.3 against 229 and 141.

This was found by iegm-site on 2026-11-02 by recomputing from the published pages rather than from the source, which is the only way it could have been found. It is the defect this note exists to prevent, present in the note itself.

One judgement call, stated because it is where the figure is most open to challenge. PMCPS-M section 202, Application, is excluded as a scope-and-effect section, and it carries two real duties: to make and retain a classification determination, and to file it where PMCPS-C section 105(d) requires. Counting section 202 would add two to the PMCPS-M primary figure. The Institute excludes it because scope sections are excluded as a class, and records the cost of that choice here rather than leaving it to be discovered.

3. When a lead-in counts, and the correction that produced this rule

A lead-in is excluded from the secondary figure only where at least one of its own items also carries an obligation term. Where the lead-in is the sole bearer of the term, it is counted.

The reason is that excluding lead-ins exists to prevent double counting, not to discount duties. Where both the lead-in and its items carry terms, counting both counts one duty twice. Where only the lead-in carries one, excluding it counts that duty zero times.

PMCPS-M section 306(a) is the clearest case. It reads "A covered operator shall not advertise in a manner that:" over four items, none of which contains an obligation term. The prohibition is the lead-in. An unconditional lead-in exclusion removes the advertising prohibition from the count entirely. Sections 302(a), 304(a) and 402(b) have the same shape: the duty is stated once, at the top, and the items beneath enumerate what it applies to.

This rule is a correction. Until 2026-10-31 the Institute applied the exclusion unconditionally to any lead-in carrying a paragraph marker, and conditionally only to unmarked ones. That was wrong, and it was wrong in both families. The figures it produced are withdrawn. They are named in section 5 rather than removed, because a withdrawn figure that is quoted somewhere needs to be findable.

4. The figures

Measured from the published source of each version with the counter named in section 6.

DocumentVersionPrimarySecondary
PMCPS-Cv1.11211
PMCPS-Mv1.1144100
Familyv1.1156111
PMCPS-Cv1.21914
PMCPS-Mv1.2186117
Familyv1.2205131
PMCPS-Cv1.32717
PMCPS-Mv1.3202124
Familyv1.3229141

PMCPS v1.3 is the current published version, issued 2026-11-01. No figure is stated here for any version that is not published, because a reader could not recompute it.

The figure that was open at v1.0 is settled. v1.0 recorded that 131 and 132 were both in circulation for the v1.2 family secondary and that neither should be treated as settled. It is 131. The 132 came from recomputing the published HTML while stripping blockquotes, which the published HTML does not contain, so the drafting note in PMCPS-C section 104 was counted. Both figures were arrived at honestly from the same definition, and the definition was the thing at fault: see the aside rule in section 2.

4a. What changed in version 1.1 of this note

The exclusion for advisory notes is now stated in terms of the published HTML, as aside elements, rather than in terms of the markdown source. The v1.3 figures are added. The open figure recorded at v1.0 is resolved in favour of 131. No rule changed: the same paragraphs were counted before and after, and v1.0's printed figures stand. What changed is that the method as written now produces them.

5. Figures withdrawn

WithdrawnWas stated asCorrected toWhy
PMCPS v1.1 secondary94, later 95111Lead-in exclusion applied unconditionally
PMCPS v1.2 secondary113, later 114131Lead-in exclusion applied unconditionally

No primary figure has been withdrawn. 156 and 205 have been stable throughout and are unaffected by the lead-in rule, which touches only the secondary figure.

6. How to check this

The counter is runner/obligations.py in the Institute's committee software, sha256 beginning d429f6fe0ce47ad8. It is not published, so the definition above is written to be implementable without it: the terms in section 1, the sections in section 2, and the lead-in rule in section 3 are the whole method.

A recomputation that disagrees with a figure in section 4 is more likely to have found a defect than to be mistaken. Two of the Institute's own figures were wrong for a month while being arithmetically sound, stable, and matched by a second implementation. That agreement was worth nothing, because the second implementation was built from the same written definition and could not have disagreed. Two instruments sharing a definition are not two instruments agreeing. If a recomputation differs, state the difference rather than reconciling it.


License and naming

(c) 2026 Aevora Solutions LLC. Published by the Institute for Emerging Gaming Markets.

This note is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). You may copy, redistribute, adapt and build upon this material for any purpose, including commercially, provided you give appropriate credit to the Institute for Emerging Gaming Markets, provide a link to the license, and indicate whether changes were made. https://creativecommons.org/licenses/by/4.0/

PMCPS and Institute for Emerging Gaming Markets identify the source of this document. A modified version may not be described as a note of the Institute. This note is not a standard and conformance with it cannot be claimed.


Institute for Emerging Gaming Markets

A standards body under Aevora Solutions LLC

Charleston, South Carolina ยท iegm.org

PMCPS obligation counting note, Version 1.0, 2026-11-01. This note is not a standard and has not been adopted by any regulatory authority.