But reformers might have overestimated the knowledge and confidence of many non-specialist primary teachers whose academic strengths were not in maths.
"Maybe we have to admit that we needed to provide more structured support for teachers than we did.
"In the ideal world, everyone would be able to run with this. Maybe they can't.
"I thought 50 per cent of the teachers were doing a fantastic job ... We just needed to have the other back-up plan, maybe, something more structured for those that it didn't work so well for."
New Zealand 9-year-olds finished bottom equal among developed countries in the 2011 Trends in International Mathematics and Science Study published in December.
Half the students could not add 218 and 191.
Education Minister Hekia Parata described the results as "extremely concerning".
Ministry of Education figures show the number of Year 8 (12-year-old) children who could answer a series of simple multiplication questions correctly within four seconds dropped from 47 per cent in 2001 - the year new maths teaching methods began - to 37 per cent in 2009. The methods were part of a ministry-led initiative called the Numeracy Development Project, which stemmed from a Government response to New Zealand's poor results in the 1995 test.
Professor Wright co-ordinated the project from 2000 to 2010.
He said it was frustrating to see New Zealand's continued poor performance, particularly in simple addition, subtraction, multiplication and division, known as basic facts.
He acknowledged that national testing showed Year 4 (8-year-old) children were too slow because they did not have the answers at their fingertips.
"They were trying to work out basic facts that you hope at that point children just knew or could derive very quickly."
Professor Wright said New Zealand had a longstanding difficulty finding teachers who were confident and knowledgeable teaching maths. Part of the problem was the low entry standard for teacher training and the limited time trainees had to study maths education, compounded by the shift from four to three-year teaching degrees.
But most of the struggling teachers were those already in the workforce, suggesting the real answer was better on-the-job training - like a maths version of Reading Recovery, which upskilled teachers as well as students.
He said New Zealand could follow Britain and the US, which had narrowed their focus and improved their results in the Trends in International Mathematics test, but it risked losing possible gains in mental calculation and problem-solving skills.
It was also a myth that highly successful countries such as Singapore and South Korea spent lots of class time on basic facts.
"There's a huge industry in those countries of after-school programmes for students. In Korea $60 billion a year is spent by parents in after-school programmes, and a large proportion of that is mathematics.
"In New Zealand we'd rather have our kids throwing netballs and kicking footy balls and bowling cricket balls, than spending their time on after-school classes, as they would in Korea.
"We need to decide what we want, I think."
The problem
NZ children have scored badly at basic arithmetic in a series of international and national tests. In one test, half our 9-year-olds could not add 218 and 191.
Some critics claim new teaching methods, which emphasise problem-solving strategies instead of traditional calculations, have confused teachers and children.
• The man who led the changes claims the system can work well, but many teachers don't have enough maths knowledge to use it properly. Many parents and teachers have debated the issue online, including at a "Bring back column addition" Facebook page at http://tiny.cc/1xuytw.
