top of page

How To Fight A War (or in this case - run a school)

These are the notes I've made while reading Mike Martin's excellent "How To Fight A War". While it looks at how effective military organizations are run - I think the same principles can (and should) easily be applied to the world of teaching. There is a huge amount that we do wrong in teaching - generations of failed students and the teaching profession having one of the highest staff turnover rates in the world - would indicate there is.

These notes will be useful for me later on - and hopefully, there are some good ideas in there worth sharing now.

The ultimate message is this - we are losing the logistical war, not the pedagogical one.


No amount of fancy CPD will make up for a teacher who hasn't eaten or slept properly, it won't compensate for a long photocopier queue or a timetable that sees them teach 7 back to back classes.

Head teachers should all teach.

Logistics: soldiers don't run backwards and forwards to collect their ammo, baristas don't run to the store room to collect coffee beans after every customer. All photocopying required should be submitted, done and then ready in each teacher's classroom each morning. The reprographics team should be large, well funded and work shift patterns to achieve this.

3 period day best facilitates planning - and prevents cognitive overload of both students and teachers - not to mention the lost learning time and potential behavior problems from shuffling around the corridors 7 times a day.

Morale: discount schemes with local businesses, practical assistance (receive their post/wash cars/collect laundry), funding for social events, family events. Free fruit and veg to allow them to hit healthy limits, free gym access.

Well funded and exciting trips - with leadership of them dispersed to give people trust buy in and identity.

A free coffee? Family days - moving house days -

Basic Discipline (enforced by visible and helpful SLT in corridors).

Greet students by door, they enter in silence.

Consistent 3 strikes policy and a staffed removal room (who also process the paperwork for detentions/messages home etc).

Weekly Homework booklets - submitted online and marked by AI where at all possible - grades instantly shared with parents in a weekly homework report (and gamify with points).

Combined Arms Operation.

What about the teaching equivalent of Combined Arms Ops?

Teachers without TA support (in primary) do not have the resources to teach. TAs without teacher support are useless - and when they intervene blind all they do is deprive students of the attentions of the teacher (or worse, do the kid's work for them). The pastoral team need accurate behaviour data reported by teachers so they know who to target, and teachers need the pastoral team to remove/rehabilitate disruptive students and provide guidance - and the SENCO team have much the same relationship.

Structure of the curriculum:

Momentum, just like for an army, is key to the psychology of learning. Small tasks, easily achievable, gaining student engagement and confidence - building up to bigger pictures and tasks.

Shaping operations are about forcing the adversary to fight on the territory and in the manner you want to them. In teaching we still expect teachers in the classroom to be a master of everything.

Why don't schools have separate homework departments? (This could be easily automated) - with homework setting, marking and feedback being given by someone other than the classroom teacher? Trying to get a teacher to do all this admin is a bit like asking an artillery Regiment to make a frontal assault on a trench with fixed bayonets - it isn't what their designed to and isn't the best use of their skills.

In a similar vain, why do teachers deal directly with parents? Surely this is yet another skill that is best delivered by a specialist, allowing the teachers to teach. Heads of years are an obvious choice, but that's not to say others couldn't be hired or step in. The guys designing new breakfast cereal are not also the same guys on the phones in the customer care base - feedback, diluted and actionable- gets through to the designers, but they are allowed to be distracted from applying their key skill in the most effective manner.

In a similar vain - attendance. This is a process that takes up significant time in a classroom (even more of you're in a school that insists on having a ridiculous number of periods a day).

This can and must he automated using technology. It is a no brainer and easily achievable. Whether that's through entry ID cards, EFID, or even something like facial recognition.

Data is something that makes teachers moan, and usually comes top of a list of things that are wrong with the profession. But. This is because, yet again, the teachers are the ones being made to do the data entry, and in the most part, the analysis themselves. Ronaldo doesn't spent the three hours after a football game inputting all the game stats into a database and writing up a report on lessons to he learned. Data analysts do that. They specialise. And it works.

Again, why are we retarding teachers by expecting them to be 1000 other things on top of just teaching? The problem is structural. It isn't because teachers are using the wrong teaching techniques or being too woke or the behaviour system is poor - the problem is that every single one of these areas isn't allowed to succeed because the teachers are the ones being expected to so them all. It is simply impossible for any modern school to succeed given the current structural model.

Another military book - Black Ops by Mark Urban. Talks about the use of intelligence by special forces during the Iraqi "surge". Four teams working in pairs and rotation, one team of special ops raiding a location and taking prisoners/seizing computers while their intel team slept - once the team returned the prisoners, they swapped over, the intel team analysing the data/questioning the prisoners and preparing a target brief ready for the next evening.

The ultimate goal needs to be to create an AI driven assessment machine at the heart of the school. Drawing in written Councillor reports, teacher observations, regular formative assessments to advise on daily seating plans, lesson plans and generate supportive resources for specific students. This linking resources (using tags and fine turning based in teacher feedback) to the data process. So when a teacher comes into work not only are the resources printed - but the plans are made and advice on supporting specific students printed. This isn't about deskilling teachers - it is about allowing them to actually use their skills of classroom management, engagement and on the spot modification, for the first time in a sustained and effective way.

Force ratios. In the military they say 3 - 1 is a good rule of thumb, never attack unless those conditions are met. In football too, it always about engineering situations where you have the free man - you dominate with man power - allowing you to pass the ball through. What does this force ratio look like in teaching?

Teachers are always outnumbered in the classroom, even with TAs we can never enjoy good overwhelming ratios. But it is a different game, and unless operating in a special school, the ratios will always be around 25-30/1. Studies from Korea and Japan seem to indicate, albeit in a hugely different academic culture, that class numbers don't really make much difference- which flies in the face of western thinking which believes that small class sizes in private schools are on the key advantages they have over state schools. Election pledges tend to focus on the number of kids per class, and there is a clear logic that a teacher is better able to devote one to one time with a student in a smaller class.

Assuming then that you're using a pedagogy which relies on interaction, rather didactic note taking (which may be the case in the East), ratios should be important. Studies show that around 15-20 is the ideal class size. ChatGPT lists the following in support of this:

Tennessee's Student/Teacher Achievement Ratio (STAR) study: This large-scale experiment found that students in smaller classes in early grades (K-3) performed better academically and had higher graduation rates, which suggests the potential benefits of reduced class sizes.

The Project STAR study (not to be confused with the above): Conducted in 1985, this study examined the impact of class size reduction on students' long-term academic and career outcomes. It found that students in smaller classes continued to show academic advantages even after their time in smaller classes had ended.

The Mathematica Policy Research study: This research evaluated the effects of class size reduction in the early grades on student achievement. It found positive effects on academic outcomes, particularly for minority and disadvantaged students.

A meta-analysis by Hattie: This study aggregated data from numerous class size-related research and found a small but positive effect of smaller class sizes on student achievement.

The SAGE (Student Achievement Guarantee in Education) project: This ongoing study in Wisconsin aims to reduce class sizes in grades K-3 in certain schools. Preliminary findings showed positive impacts on reading and math achievement.


I think ratios are important - in terms of student numbers but also periods in a day. How could more favourable ratios be engineered? Could you have different size classes for different types of learning? Have some lessons that are very much didactic- and have larger class sizes for those - and others that are more interactive, which have smaller. Much like a university course might involve lectures to hundreds of students following by seminars on the content involving only a handful, to beat discuss and apply their new knowledge? Does a library lesson need a qualified English teacher to oversee their class of 30? Or could lesser qualified/more specialists persons oversee a group of 60-90 in one library lesson sitting? Could technology provide the didactic lessons en-masse? 60 kids at a time completing interactive tasks via VR headsets, followed by small groups of 20 with qualified teachers discussing and applying their new learning/VR experiences? Could each subject break their tasks/learning episodes into specific styles that would in turn dictate the class ratio? Could this work within a traditional timetable?

Timing is also key. When do your students attempt the difficult questions? Have they had a chance to master the smaller skills required? This is well known and most well run departments ensure this happens, but I think you'd be surprised how often this doesn't happen. Turnover of middle management, the knee jerk inclusion of new priorities from SLT or trends in the subject. For example, it is easy to see how in the last few years a history department, quite rightly seeking to be as relevant as possible, may have included a few lessons on previous pandemics, then a few lessons on black history and then a few more on Ukraine and before you know it a beautifully curated curriculum which steadily built up skills and revisited them has gone out the window. It would take a very vigilant head of department to ensure curriculums are kept up to date but simultaneously maintain the synchronisation of the skills you want students to develop.

One of the most poorly executed areas of education is in the transition between primary and secondary school.


Featured Posts
No posts published in this language yet
Once posts are published, you’ll see them here.
Recent Posts
Search By Tags
Follow Us
  • Facebook Basic Square
  • Twitter Basic Square
  • Google+ Basic Square
bottom of page