Corrections

This page lists every correction of substance to a figure or a finding on the site, with the date it was made, what the page said before, what it says now and why. A change of wording that leaves the meaning as it was is not listed. To report an error, use the feedback form or write to info@pinpointventures.in. The terms page sets out how a correction is handled.

13 October 2026

Uttar Pradesh

• The Maharashtra pages said the Chief Electoral Officer had published no count by reason and nothing by constituency beyond a list of 38 seats it chose, and they gave the reasons for the state as the press reported them, in lakh. The Chief Electoral Officer's tables of 1 September 2026 give the five reasons for every district and constituency, and a statement enclosed with its press note of 16 September 2026 gives the roll before and the names left off for all 288 constituencies. The pages now use them. The state's reasons were 34.81 lakh dead, 154.55 lakh shifted, absent or other and 17.63 lakh enrolled at more than one place, rounded and 10,513 over the total. They are now 34,82,231, 1,54,39,754 and 17,66,502, which add up to the total exactly. The All India pages use the exact figures too.
• The Uttar Pradesh page said no split by sex of the names left off had been published. The Chief Electoral Officer gave one to the Hindustan Times on 18 January 2026, 1.55 crore women and 1.34 crore men, rounded to the lakh, and the All India women page already used it. The page now shows it. Women were 53.6% of the names left off and 46.8% of the roll before the revision, 21.5 women left off for every 100 on the roll against 16.3 men.

Uttarakhand

• The Uttarakhand pages said the names left off the draft had been published for the state only. The Chief Electoral Officer published a table of the draft by constituency on 14 July 2026, which the site had missed. It gives the names left off in each of the 70 constituencies, and they add up to the state total. The pages now show them, and the map is shaded by them. Kichha lost 18.2 of every 100 names on its roll before the revision and Didihat 4.2.
• The district factors, border and urban pages measured Uttarakhand by the names removed after the draft. They now use the names left off the draft, the same measure as most other states. On the factors page the reading for clean cooking fuel is unchanged in direction, and Uttarakhand joins the urban page. On the border page, Uttarakhand's border districts lost slightly more than the rest of the state on both measures.

Tamil Nadu

• The Tamil Nadu elections page worked out turnout as the votes over the final roll of 23 February 2026. The roll the election was held on, that of 6 April 2026, was published by the Chief Electoral Officer and is larger in every constituency, so turnout is now worked out on it: 86.0% for the state, where the page said 87.0%. Weighed at that turnout, the living names removed outnumber the margin in 168 seats, where the page said 169.
• The Tamil Nadu page said no roll by sex from before the revision had been published. The Chief Electoral Officer published the final roll of 6 January 2025 by constituency and sex. The women section now sets the fall from it to the draft against it.

10 October 2026

These followed a check of every page against its sources, and a re-reading of the news of 7 to 10 October.

Bihar

• The comparison with district income per head is back on the Districts page, from a published source. Income per head is now per capita gross district domestic product for 2022-23 at constant 2011-12 prices, from Table A 1.7 of the Bihar Economic Survey 2024-25. Districts with more income per head lost slightly fewer names, with a rank correlation of −0.23, and income moves closely with the urban share and with women's literacy, so the three cannot be told apart.
• The Districts page compared removals with each district's income per head and found no relationship. The income figures had no source and several looked like placeholders, so the comparison has been taken down. No other figure replaces it.
• Surnames in the first caste classification were checked against the official lists for Bihar compiled by Jolad, Kalyani and Baharadia (Harvard Dataverse, doi:10.7910/DVN/WT5VIY). Turi is now Scheduled Caste, where it had been Extremely Backward Class. Dom and Bhuiyan are now Scheduled Caste, where they had been Scheduled Tribe. Dhobi is Scheduled Caste for Hindus and a backward class for Muslims, so the surname now counts for both groups. Under that classification the Scheduled Caste share of the names removed rose from 19.8% to 20.0% and the Scheduled Tribe share fell from 1.3% to 1.1%. No reading on the People page changed.
• The report gave 9.25 removals for every 100 electors on the final roll. The figure is per 100 electors on the 2025 election roll of 7,46,91,135.
• The dashboard and the explorer said 3,418 booths lost more than 200 names each. The 3,418 booths lost 200 or more.
• The typical booth lost 72 names. The dashboard and the report had said 71, the middle of the constituencies' middle figures.
• Gopalganj was described as the district that lost the most names. It lost the most for its size. Patna lost the most names, 4,15,580.
• On the Elections page, the NDA's lead over INDIA, once each seat's 2020 result is allowed for, grew by 4.0 to 5.2 points in each third of the seats ranked by removals. It had said 4.5 to 4.7. Rates had been stored to two decimals, and ties at the rounded value put a seat in the wrong third. The model's slope did not change.

All India

• On The SIR in Numbers, Karnataka's Assembly seats where the names left off, spread by electors, exceed the 2023 margin are now 208 of 224. The page had said 207. Karnataka's results had been taken from a published compilation, and they are now the Election Commission's statistical report. Five margins differ slightly between the two, and the electors differ in every seat because the report counts service voters. With the report's electors as weights, Chitradurga's share of the names left off comes to just over its margin of 53,300. Puducherry's figures now come from the same report, with no change to its count, and Nagaland, which had been left out for want of a source beyond Wikipedia, is now included.
• Uttarakhand's notices were counted at 24,30,000, the number the state set out to serve. The number served was 20,27,000. The notices sent across the country are now 21.1 for every 100 electors on the draft rolls, where they had been 21.2.
• The process page said six of the phase 3 states had their schedules moved. It was seven of 19, namely Andhra Pradesh, Haryana, Karnataka, Maharashtra, Delhi, Meghalaya and Mizoram. The count is now worked out from the schedules.
• Chhattisgarh's final roll is now shown with 3,43,801 names added and 1,08,807 removed after the draft, as the Chief Electoral Officer published them. The site had shown only the net rise. Puducherry's additions after the draft are now 42,719.
• The comparisons page and the brief said the final rolls made up at most 14.1% of the names left off. That is the figure for the 21 states taken together, and it now says so. Leaving out the 60 lakh names West Bengal held for adjudication, it is 22.0%. Taken one by one the states range from 85.6% in Lakshadweep to below zero in five.
• The cities page said that in six of seven cities the rate fell from the city to its suburbs and again to the rest of the state. The first step holds in six of seven. Both steps hold in three of six, and Delhi, a city and a state at once, is left out of that count.
• The timeline entry of 20 February 2026 joined two orders. The judicial officers from Calcutta High Court were requisitioned by the Supreme Court's order of 20 February, and those from Jharkhand and Odisha under its order of 24 February. They are now two entries.
• The Supreme Court's judgment in Mihir Rajesh Shah is now cited as Criminal Appeal No. 2195 of 2025, (2026) 1 SCC 500. The neutral citation given before is disputed between sources.
• Press reports on the process page named eight private people, among them voters, booth level officers and people who died during the revision. They are now described by their role.

Hyderabad and Telangana

• On the hearings page, in the seat with the busiest single slot, the typical slot listed 50 cases. The page had said 38.
• The third most common reason for a notice is Progeny of six or more, 15.1% of the flags. The page had given a different rule as third, because the rules were not sorted by count.
• The Telangana page said 18,12,367 electors sent a notice had not responded. That figure is the cases not yet decided, which takes in electors who were heard and are awaiting an order. It now says so.

West Bengal and other states

• West Bengal's 1,607 names restored by the tribunals in time to vote is a press figure. It had been presented beside the Commission's figures as if it were one of them.
• Rates had been stored to two decimals and shown to one, which rounded some figures the wrong way. Among those corrected, Dahod in Gujarat is 10.6 per 100 (it showed 10.7), Lalitpur in Uttar Pradesh is 10.0 (9.9), the share of West Bengal's removals entered as shifted is 34.2% (34.1%), and part 22 in Hyderabad is 80.3% (80.2%). Every rate is now stored to four decimals.